Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Networks of blood proteins in the neuroimmunology of schizophrenia

Networks of blood proteins in the neuroimmunology of schizophrenia Levels of certain circulating cytokines and related immune system molecules are consistently altered in schizophrenia and related disorders. In addition to absolute analyte levels, we sought analytes in correlation networks that could be prognostic. We analyzed baseline blood plasma samples with a Luminex platform from 72 subjects meeting criteria for a psychosis clinical high-risk syndrome; 32 subjects converted to a diagnosis of psychotic disorder within two years while 40 other subjects did not. Another comparison group included 35 unaffected subjects. Assays of 141 analytes passed early quality control. We then used an unweighted co-expression network analysis to identify highly correlated modules in each group. Overall, there was a striking loss of network complexity going from unaffected subjects to nonconverters and thence to converters (applying standard, graph-theoretic metrics). Graph differences were largely driven by proteins regulating tissue remodeling (e.g. blood-brain barrier). In more detail, certain sets of antithetical proteins were highly correlated in unaffected subjects (e.g. SERPINE1 vs MMP9), as expected in homeostasis. However, for particular protein pairs this trend was reversed in converters (e.g. SERPINE1 vs TIMP1, being synthetical inhibitors of remodeling of extracellular matrix and vasculature). Thus, some correlation signals strongly predict impending conversion to a psychotic disorder and directly suggest pharmaceutical targets. Introduction endothelial cell function and general integrity. Such Circulating levels of immune system proteins and rela- findings support the hypothesis that psychosis involves ted signaling agents are consistently altered in schizo- brain dysregulation by an altered peripheral immune phrenia. This observation includes unmedicated first system and aberrant signaling at BBB. 1,2 episode psychosis patients and persons at clinical high- Persons meeting clinical high-risk criteria have about a 3–5 risk who subsequently convert to psychosis . Consistent 20% risk of developing a psychotic disorder within two findings include proteins in the immune system acute years, and that is 100-fold higher than the 0.2% general phase response and in the plasminogen activating sys- population risk . Baseline factors differentiating clinical tem . Many of the proteins influence brain function high-risk subjects that convert to psychosis vs do not directly, crossing blood–brain barrier (BBB) and signaling convert may be of etiologic significance. Previous psy- glia or other perivascular cells. These immune signaling chosis risk prediction studies compared levels of immune molecules also regulate brain function by influencing BBB signaling molecules, individually or in a linear combina- 3–5,7 tion . Since responses of the peripheral immune sys- tem are highly coordinated, investigation of the correlation patterns of immune signaling molecules might Correspondence: Clark D Jeffries ([email protected]) Renaissance Computing Institute, University of North Carolina, Chapel Hill, NC, also be informative. In this analysis we sought networks of USA highly correlated immune molecules in persons at clinical Department of Psychiatry, University of North Carolina, Chapel Hill, NC, USA high-risk who developed psychosis, compared to those Full list of author information is available at the end of the article Larry J. Seidman deceased on 7 September 2017. © The Author(s) 2018 Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to theCreativeCommons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/. 1234567890():,; 1234567890():,; 1234567890():,; 1234567890():,; Jeffries et al. Translational Psychiatry (2018) 8:112 Page 2 of 10 who did not develop psychosis over the two-year follow- A second key concept is the pleiotropic roles of proteins up period. We also analyzed data from unaffected com- customarily considered in the context of hemostasis vs parison subjects. In some respects, the correlation net- hemorrhage. Specifically, as Nave and Ehrenreich demon- works were strikingly different in converters. Some strated , “It is becoming apparent that coagulation factors distinguishing proteins were modulators of extracellular do much more than simply act in the blood-coagulation matrix (ECM) components and BBB. cascade.” For example, fibrinogen associates with schizo- 18 19 20 Schizophrenia pervasively impacts brain functions, phrenia , Alzheimer disease , and multiple sclerosis (MS) . typically causing in adolescence or early adulthood dis- A third key concept is that the peripheral immune ability that is chronic and relapsing. Improved clinical system impacts physiological and pathological brain 21–23 outcomes are often associated identification and treat- function . We have hypothesized that peripheral ment of the disorder early in its course , implying the immune system dysregulation may contribute to devel- desirability of seeking reliable predictors based upon opment of psychosis. Furthermore, the construction of readily accessible biomarkers. Moreover, early indicators the above classifier survived permutation testing very might be close to the ultimate causes of schizophrenia. well, implying that certain peripheral blood plasma pro- Thus, researchers (e.g., Domenici et al. and Dickerson teins and conversion of patients to schizophrenia are et al. ) have investigated in particular proteins and other associated. Finally, dysregulation of BBB permeability, blood plasma analytes that distinguish unaffected com- transendothelial cell migration, or remodeling of brain parison subjects from patients with schizophrenia or ECM or vasculature—all impacted by the immune system 11,12 subsets of such patients . Included among the analytes —could obviously be important in the events leading to have been concentrations of circulating cytokines and frank psychotic mood disorders including schizophrenia. other immune system signaling molecules. These can be altered in schizophrenia across all stages of the disorder Materials and methods including the prodromal stage. Subjects One key concept of this paper is information from The aims and methods of North American Prodrome networks. As Fredrickson et al. stated, “…accumulation Longitudinal Study (NAPLS2) were described in detail of many individually noisy indicator variables can yield previously . Briefly, NAPLS2 is an eight-site observa- highly stable estimates of the underlying factors they tional study of predictors and mechanisms of conversion share in common.” Thus, we sought networks of highly to psychosis. The NAPLS2 cohort includes 765 subjects at correlated signals among each of three groups (converters, clinical high-risk for psychosis, based on the Criteria of nonconverters, and unaffected comparison subjects) from Prodromal States, determined by the Structured Interview our North American Prodrome Longitudinal Study for Prodromal Syndromes, and rated with the Scale of 14 25 (NAPLS) project . The emphasis was analysis of data Prodromal Symptoms (see Supplement). In addition, the collected at the prodrome state (see Supplement Figure cohort includes 280 demographically similar unaffected S1). To place the present work in context, NAPLS is a comparison subjects. Subjects were between ages 12 and multi-site program that has accumulated, cleaned, stored, 35 at baseline. Psychiatric diagnoses were determined by analyzed, and reported many types of clinical and the Structured Clinical Interview for DSM IV . Clinical laboratory assays, leading to proposals of various pre- assessments were performed every six months, and sub- dictors and mechanisms for the development of jects were followed for up to two years. The study was psychosis . approved by the Institutional Review Board at each site, Thus, the present emphasis differs from some earlier and each subject provided written informed consent or works in that it is longitudinal and it pertains entirely to assent, with a parent or guardian also consenting for networks. The three networks from the three groups are minor subjects. Demographics for subjects in the present formed by proteins that are highly correlated over subjects study are provided in Table 1. in each group, far more highly correlated than could be reasonably explained by chance. All of the reported cor- Plasma analytes relations are positive because there we observed no During a baseline visit, blood samples were drawn using negative correlations of the same, very high magnitude. Becton Dickenson P100 blood collection tubes containing Many of the distinguishing analytes we found turned EDTA (as anticoagulant), proprietary protein stabilizers, out to be prominent in the immune system and its and a mechanical separator. Mean processing time to interaction with trophic factors and ECM components. −80 °C storage was 28 min (SD = 2 min). Plasma samples Thus, the proper arena of this paper became neu- were subsequently sent on dry ice to Myriad RBM (Austin roimmunology. Particularly in the last decade, many TX), a laboratory that has maintained CLIA Accreditation important reports have developed this view of mental since 2006. Samples were analyzed with the Human illnesses (e.g., Khandaker et al. ). DiscoveryMAP v. 1.0 assays of 185 analytes associated Jeffries et al. Translational Psychiatry (2018) 8:112 Page 3 of 10 Table 1 Demographic and clinical characteristics of study comparison subjects. Quality control, normalization subjects taken at baseline of longitudinal study methods, and tests with duplicated samples were per- formed, as previously described . Unaffected Clinical high-risk, Clinical high- comparison nonconverters risk, Data analyses (UC) N = 35 (CHR-NC) N = 40 Converters We used an unweighted co-expression network analysis (CHR-C) N = to identify highly correlated networks of analytes in each of the three groups. Pearson correlations were calculated using macros and built-in functions in Excel. Limited by Age, average (SD) 20 (4.5) 19.5 (4.6) 19.2 (3.7) the smallest group size (converters, n = 32), we performed Ancestry random re-sampling with replacement 10,000 times by %Caucasian 60%, 65% 55% drawing subgroups of 28 subjects from each of the three %African 31% 17.5% 21% groups. In other words, the 9870 pairs of analytes were compared in 28-dimensional space 10,000 times to cal- %Asian 9% 17.5% 24% culate Pearson correlations. We compared those correla- Sex, % female 34% 37.5% 30.3% tion values to a common threshold and then recorded the SES, average (SD) 4.8 (1.8) 4.5 (2.3) 4.5 (1.8) number of times a pair exceeded the threshold (flowcharts Time blood draw, 12.:12 pm 12:39 pm (2.0 h) 11:59 am in Supplement Figure S2). The correlation threshold average (SD) (1.85 h) (1.79 h) selected was 0.7662; if two 28-dimensional vectors were populated by a Gaussian distribution, then the probability Prescription medication of their correlation exceeding this value would be 1E-6. Antipsychotic 0% 25% 13% Thus, with 9870 pairs of analytes, we expect only 1 such Antidepressant 1% 30% 25% value among all pairs about once in 100 trials. Stimulant 0% 8% 6% Considering all 107 subjects and all 9870 pairs of ana- lytes at once, distribution analysis using EasyFit Mood stabilizer 0% 5% 3% (MathWave, Dnepropetrovsk, Ukraine) revealed that no Benzodiazepine 0% 5% 13% normal distribution fit the observations but the Johnson NSAID 0% 0% 0% S distribution did. Technically, normal distributions Antibiotic 0% 0% 0% failed to fit the observed distribution of correlations in both Kolmogorov–Smirnov and Anderson–Darling tests Substance Use for even α = 0.01; by contrast, the Johnson S distribution Tobacco use 9% 30% 44% achieves α = 0.2 in both tests. The Johnson S distribu- Alcohol use 46% 48% 38% tion accommodates a one-sided tail of very strong, posi- Marijuana use 9% 25% 31% tive correlations not balanced by even one negative correlation of equal magnitude. Therefore, we focused on Current co-morbid DSM IV Diagnosis only the very strong, positive correlations (Fig. 1). Depression 0% 45% 50% We validated our approach by first randomly permuting Anxiety 3% 60% 56% all 107 subjects, then analyzing 28-subject subsets of the g,* Disorders pseudo-groups exactly as if they were from the true data. Neither the design nor the implementation of the algo- CHR-C vs UC FET p-value = 0.047, CHR-NC vs UC Fisher Exact Test (FET) p-value = 0.001 rithm conferred any obvious, consistent distinction to CHR-C vs UC FET p-value = 0.011, CHR-NC vs UC FET p-value = 0.002 edge counts of graphs of very strong correlations in the CHR-C vs UC FET p-value = 0.047 CHR-C vs UC FET p-value = 0.001, CHR-NC vs UC FET p-value = 0.02 three pseudo-groups. This result indicates that both the CHR-C vs UC FET p-value = 0.020, CHR-NC vs UC FET p-value = 0.056 design and implementation of the program correctly CHR-C vs UC FET p-value < 0.0001, CHR-NC vs UC FET p-value < 0.0001 CHR-C vs UC FET p-value < 0.0001, CHR-NC vs UC FET p-value < 0.0001 achieved unbiased identification of correlation networks. *Depression disorders include Major Depression, Depressive Disorder Not Measures of network complexity include those of Otherwise Specified, and Dysthymic Disorder. Anxiety Disorders include Obsessive Compulsive Disorder, Post-Traumatic Stress Disorder, Panic Disorder, Bonchev and Buck (see their formulas 19a,b) in which Agoraphobia, Social Phobia, Specific Phobia, Generalized Anxiety Disorder the normalized edge complexity is the number of (undirected) edges e divided by the maximum possible with immune system function, hormonal responses, oxi- number of edges using the same number of vertices v, dative stress, and metabolism. In total 141 analytes passed namely, v*(v-1)/2. preliminary quality control and were used in our analyses. Protein network analyses were conducted using Inge- We standardized to z-scores the results for each analyte nuity Pathway Analysis (IPA) (QIAGEN N.V., Venlo, The using the average and SD values of the unaffected Netherlands). Jeffries et al. Translational Psychiatry (2018) 8:112 Page 4 of 10 Results draws of 28 subjects from each group are shown for Protein correlation networks distinguish the three groups unaffected subjects (Fig. 2), nonconverters (Fig. 3), and There was a striking loss of network complexity going converters (Fig. 4). from unaffected to nonconverters and thence to converter For unaffected, nonconverter, and converter groups we subjects. Analyte pairs with Pearson correlations exceed- calculated a simple metric of graph complexity, namely ing an a priori threshold in at least 5000 of 10,000 random the ratio of numbers of edges to vertices; respective values were 1.48, 1.30, and 0.89 edges per vertex. Normalized edge complexities were 0.134, 0.119, and 0.062. Moreover, it is visually apparent that the three graphs are quite dif- ferent in terms of edge densities. Additional graphs and experiments with alternative choices of thresholds led to the same trend and are in the Supplement. This trend toward simplicity going from unaffected to nonconverter to converter for plasma proteins parallels the same for leukocytic microRNAs reported earlier . Notably, in unaffected subjects there was a network involving the proteins SERPINE1 (plasminogen activator inhibitor-1), MMP7, MMP9 (activated), MMP9t (total), and MMP10 (matrix metalloproteinases). While SER- PINE1 is anti-remodeling, the others are pro-remodeling, and a balance (correlation of some pairs) might be expected in hemostasis vs hemorrhage. In contrast, non- converters and converters lacked these strong SERPINE1 correlations. Furthermore, while the important inhibitor of remodeling TIMP1 (tissue inhibitor of metalloprotei- nases) was absent (at the chosen threshold for inclusion) Fig. 1 A histogram with 20 bins of 9870 correlation values from strong correlations in unaffected and nonconverter among 141 analytes over all 107 subjects. The shown distribution is a Johnson S fit with four parameters: gamma= −1.2967, delta= graphs, it generally exceeded the threshold correlation 2.2624, lambda= 0.26593, xi= −0.12371. The present study is value with SERPINE1 in converter subjects. In detail, the distinguished from many others by focusing on the tail of very strong, SERPINE1 vs TIMP1 correlations exceeded the threshold positive correlations (blue box) that are not balanced by any negative in the three sets of 10,000 random draws of 28 subjects in correlations of the same magnitude totals of <10, <10, and >8300 times, respectively. As sets, Fig. 2 Unaffected comparison subject data yielded a graph of strongly correlated analytes with 23 analytes (vertices) and 34 robust correlations (edges). Blood plasma proteins are labeled by their gene common symbols. The correlations in orange appear in all three graphs. We note eight analytes that include SERPINE1, and in particular SERPINE1 correlations include the matrix metalloproteinases MMP7, MMP9, and MMP10. MMP9t denotes an assay for both pro-MMP9 and mature MMP9 Jeffries et al. Translational Psychiatry (2018) 8:112 Page 5 of 10 Fig. 3 Nonconverter data yielded a graph of strongly correlated analytes with 23 analytes and 30 robust correlations. SERPINE1 correlations are completely absent, suggesting a loss of requlation of expression of the gene Fig. 4 Converter data yielded a graph of strongly correlated analytes with 27 analytes and 24 robust correlations. Four SERPINE1 correlations are present, suggesting changes in requlation of the SERPINE1 gene compared to unaffected and nonconverter assays. Remarkable is the gained correlation of SERPINE1 and TIMP1 because both proteins inhibit anticogulation and vascular remodeling in some contexts and both generally promote anti-inflammation. Furthermore, TIMP1 is completely absent in unaffected and nonconverter graphs. Also, the strong SERPINE1 correlations with matrix metalloproteinases (MMPs) in the unaffected graph are absent in this converter graph the actual SERPINE1 and TIMP1 correlation values over is known to promote mutual expression of SERPINE1 and 10,000 draws for converter subjects (mean correlation = TIMP1 and so might be part of the explanation (see 0.79, SD = 0.03) tended to be significantly greater (p- Discussion). value < 1E-100) than the values in nonconverters (mean From a previous study by Domenici el al. we evaluated = 0.55, SD = 0.06) and unaffected subjects (mean = 0.50, SERPINE1 vs TIMP1 correlations. The subjects con- SD = 0.09). The dramatic distinction of SERPINE1 and tributing to that study were schizophrenia patients and TIMP1 correlations in converters is further illustrated in controls. We calculated correlations for 10,000 random the Supplement (Figure S10). subsets of 200 subjects drawn randomly from 267 control What could explain the extremely high correlations of subjects and 229 schizophrenia patients. The results SERPINE1 and TIMP1 in converters only? We note that (Supplement Figure S11) indicated patients had far another protein, CTGF (connective tissue growth factor), stronger SERPINE1 vs TIMP1 correlations than controls. Jeffries et al. Translational Psychiatry (2018) 8:112 Page 6 of 10 Subnetworks common among all three groups psychosis cannot be simple. However, Hoirisch-Clapauch 34,46 Several pairs of analytes were highly correlated in all et al. . provided additional reasoning and evidence of a three groups (highlighted in orange in Figs. 2, 3, and 4). connection, finding lower PLAT levels among 70 schizo- Among them was the complete graph formed by inter- phrenia patients compared with 98 age-matched controls. leukins 3, 5, 7, and 13. We examined the correlation Since the principal regulator of PLAT levels is SERPINE1, 9 46 pattern of these interleukins in the Domenici data for 267 Hoirisch-Clapauch and Nardi postulated that inflam- unaffected subjects. As shown in the Supplement Figure matory conditions could increase the risk of schizo- S9, the four interleukins were highly correlated in that phrenia through mechanisms involving SERPINE1 levels. cohort as well. PLAT is highly pleiotropic as, therefore, might be some of its regulators. Two of many reported PLAT substrates Discussion directly relevant to psychiatry have functions in epidermal Immune correlation networks and psychosis risk growth factor receptor (EGFR)-mediated neuroprotec- 47,48 Psychosis conversion in high-risk subjects was char- tion and the dynamics of the glutamate receptor and acterized by marked simplifications of networks of cor- ion channel protein NMDAR (antagonized by phencycli- 49,50 related proteins that regulate tissue remodeling; this dine in a model of schizophrenia) . Furthermore, might be consistent with the hypothesis of BBB dysre- PLAT is expressed in brain endothelial cells, neurons, and gulation in schizophrenia . Notable were graph changes microglia. It is strongly and negatively correlated with involving SERPINE1. The serine protease inhibitor SER- mRNA for OCLN (occludin) , a protein important in the PINE1 is the major inhibitor of the plasminogen activator stability and permeability of tight junctions. Reactive PLAT; PLAT activates plasminogen to plasmin (the final oxygen species can cause rapid release of PLAT in substrate of the fibrinolytic system) which in turn monocyte and endothelia cocultures . In a related effect degrades fibrin. Thus, a focus infra is SERPINE1 and (noting the above correlations in unaffected subjects), related proteins . MMP9 promotes monocyte migration through brain Relevant to the present study, schizophrenia patients are parenchymal basement membrane . subject to increased risk of cardiovascular disease, Consistent with our findings, there is also an emerging reduced risk of certain types of cancer, and possibly some literature implicating the plasminogen pathway in schi- aspects of accelerated aging; all these general observations zophrenia. This literature points towards down-regulation have in turn been associated with SERPINE1 or related of PLAT, up-regulation of factors including SERPINE1 32–36 34,37,46,53 proteins that inhibit PLAT activation . , and consequences More particularly, connection of schizophrenia with of elevated fibrinogen (fibrin) including incorrect differ- proteins important in coagulation homeostasis has been entiation of progenitors of oligodendrocytes into astro- amplified by recent studies of Hoirisch-Clapauch, Nardi, cytes in MS . This misguided differentiation might relate and their colleagues , starting with anecdotal descrip- to reported low levels of oligodendrocytes in layer five of tions of possible benefits to five schizophrenia patients of prefrontal cortex in schizophrenia . treatment with warfarin (which would seemingly counter For data from the same subjects presently studied we the effects of increased SERPINE1 and decreased PLAT previously reported that proteins involved in the plasmi- activity). They proposed that normalization of levels of nogen pathway, including MMP7 and Factor 7, predicted PLAT in the psychotic brain might enable long-term psychosis conversion . In our previous publication we did remission of psychotic symptoms. not find that expression levels of SERPINE1 or TIMP1 PLAT is secreted by many cell types including blood were substantially increased going from nonconverters to vessel endothelial cells that are known to express diverse converters. By contrast, the current study suggests that genes differentially even among regions of the same the extreme changes in correlation networks of SER- 38–40 artery . In neurons PLAT occurs in dendrites and PINE1 and related proteins, especially TIMP1, may synapses; it is stored in pre-synaptic vesicles that cross the indeed be informative. synaptic cleft after depolarization, subject to regulation by Most notable is an interpretation of CTGF in idiopathic 41,42 55,56 57–59 astrocytes . pulmonary fibrosis and other diseases that might Aside from hemostasis, PLAT has related roles in explain the remarkable correlation of SERPINE1 and 31,43 smooth muscle signaling in blood vessel tunica , sev- TIMP1 levels in converters. Unfortunately, in our blood eral types of ECM signaling , and BBB permeability in plasma assays CTGF was not detected in 50–70% of the the context of cerebral ischemia . Given the large num- subjects in all three of the groups, precluding its corre- bers of schizophrenia patients who are comorbid for lation analysis. However, regarding CTGF as a known cardiovascular diseases, who require treatment with driver of SERPINE1 and TIMP1 levels, we note that local warfarin or functionally related pharmaceuticals, and who upregulation of CTGF by a type of alveolar cell, resulting are not in remission, the intersection of hemostasis and in proliferation of local fibroblasts, has been proposed as a Jeffries et al. Translational Psychiatry (2018) 8:112 Page 7 of 10 surrogate biomarker of idiopathic pulmonary fibrosis MMP9 is also an important factor in ECM maintenance (IPF); CTGF is therefore the target of ongoing IPF and remodeling, assayed herein both as mature MMP9 clinical trials of the fully human monoclonal antibody and as the total of proMMP9 and mature MMP9 (labeled Pamrevlumab by FibroGen (San Francisco CA) (trial MMP9t in our graphs). It regulates glutamate receptors, NCT01890265 at ClinicalTrials.gov). modulates physiological and morphological synaptic The serine protease inhibitor SERPINE1 is a “suicide plasticity, and is regulated by glutamate at excitatory synapses protein” in the serpin superfamily and is the major inhi- . MMP9 and other extracellular proteases such bitor of the plasminogen activator PLAT and related as plasmin convert proBDNF to mature BDNF . PLAT, protein PLAU (plasminogen activator urokinase). Thus, PLAU, plasmin, and MMP9 all have roles in BBB dis- SERPINE1 is a crucial down regulator of fibrinolysis and ruption after stroke . MMP9 itself is activated in pre- ECM degradation . In wound healing, SERPINE1 con- sence of oxidative stress and in turn promotes the tributes to regulation of cell proliferation and tissue receptor for advanced glycation end-product (RAGE) (to remodeling , as well as other aspects of cell signaling and be described in a forthcoming paper by D. Dwir et al.), 44,45 migration . said to induce a self-reinforcing cycle of inflammatory 73,74 Studies of samples from clinical high-risk psychiatric responses and further oxidative stress . In addition, patients in the literature and in our data (Supplement MMP9 is involved in the degradation of the perineuronal Figure S10) can exhibit somewhat elevated levels of net, a type of ECM that wraps fast-spiking parvalbumin SERPINE1. Interestingly, de Fouw et al. reported that interneurons; the perineuronal net is known to be affected the glycoprotein Protein S (PROS1) accelerated neu- in the medial prefrontal brain of schizophrenia 75,76 tralization of SERPINE1, thereby increasing PLAT activ- patients . In mice, intravenously administered PLAT ity; this finding suggests possibilities for indirect control was detected within the brain parenchyma and cere- of PLAT levels and function. In its free form, PROS1 is a brospinal fluid, having crossed the BBB by transcytosis . cofactor of protein C (PROC) in the anticoagulation In summary, for converter subjects SERPINE1 was no pathway. PROS1 and PROC are vitamin K-dependent longer correlated with antithetical MMPs, but instead plasma proteins , included among the cofactors in the became highly correlated with the synthetical TIMP1. inactivation of the prothrombinase complex. PROS1 also This observation and the above integration of these pro- exists in a complex with complement C4B-binding pro- teins and CTGF in multiple disorders suggest the poten- tein . Complement inhibitor C4B-binding protein tial importance of CTGF inhibition. Several inhibitors of CTGF have been proposed and studied; for example, enhances plasminogen activation . Isotypes of the C4 protein have been implicated in schizophrenia gene insulin and an siRNA have been studied in the context of association studies . inhibition of vascular remodeling in diabetic retino- SERPINE1 levels are up-regulated by inflammatory pathy . However, certain advantages attend a mono- cytokines including IL6, TNF, and TGFB1, as reviewed by clonal antibody for CTGF inhibition. Brown et al. . These researchers postulated that inflam- matory conditions increased the risk of schizophrenia Interleukin networks common in all subjects through a combination of mechanisms involving We found that IL3, IL5, IL7, and IL13 form a highly increased SERPINE1 levels or decreased free PROS1 correlated module in all groups, and replicated this finding levels. in an external test set of unaffected subjects (Figure S8), PLAT was not measured in our assays, but more should suggesting co-regulation and closely related functions of 66,67 be said about some of its extensive functions . PLAT these cytokines. Three genes, IL3, IL5, and IL13 (and also degrades ECM and in particular cleaves the large glyco- IL4), exist as a cytokine gene cluster on chromosome arm protein RELN, all contributing to cell migration and tissue 5q. Unfortunately, IL4 detection was weak (below LLOQ) remodeling, particularly revascularization. PLAT affects or missing in more than half our samples, precluding neural, endothelial, and glial cells, driving multiple, correlation analysis with the rest of that cluster. But IL4 sometimes opposing effects by activating or otherwise and IL7 are among the six “γ-chain utilizing” interleukins, regulating diverse transcription factors and receptors. while IL3, IL4, IL5, and IL13 are the four “IL4-like” Regarding the neurotrophin BDNF (brain-derived neu- interleukins; both sets of proteins are considered tightly rotrophic factor), PLAT promotes the cleavage of neu- packed α-helices in a four-helix bundle motif of short core ronal proBDNF to mature BDNF, a function essential for helices . There are about 45 human interleukins, so the late-phase long-term potentiation . Pro- and mature members of the identified correlation cluster are more BDNF have important and opposing effects on synaptic similar than would be expected by chance. plasticity, regulation of neurogenesis, and neuronal sur- The top shared pathway for IL3, IL5, IL7, and IL13 from vival . Only the mature BDNF was assayed by us; it is application of IPA was “hematopoiesis from multipotent prominent in Figs. 2–4. stem cells” (p = 1.6E-9). IL3 promotes hematopoietic Jeffries et al. Translational Psychiatry (2018) 8:112 Page 8 of 10 stem cell proliferation, and IL7 regulates development of fibrogen.com/pamrevlumab-trials/) or related agents of the common lymphoid progenitor cells and development CTGF inhibition. The potential medical value of inhibi- of B-cell, T-cell, and NK-cell lineages. In detail, IL3 reg- tion of CTGF in various diseases has long been recog- ulates development of the common myeloid progenitor, nized. The recent appearance of Pamrevlumab as a and IL5 regulates development of neutrophils and eosi- potential treatment for IPF and its repurposing suggested nophils . Less is known regarding IL13, but its receptor, herein for prevention of conversion to psychosis might be especially important. Repurposing is a drug development IL12A1, is selectively expressed on CD62L+ cells (com- mon lymphoid progenitors), again suggesting a role in strategy with huge savings of time and expense. regulation of lymphoid cell differentiation . Normally, Author details the circulating proportions of neutrophils, lymphocytes, Renaissance Computing Institute, University of North Carolina, Chapel Hill, NC, monocytes, and eosinophils are relatively constant, so USA. Department of Psychiatry, University of North Carolina, Chapel Hill, NC, USA. Department of Psychiatry, Lausanne University Hospital (CHUV), possibly IL3, IL5, IL7, and IL13 in our graphs contribute Lausanne, Switzerland. Laboratory of Neurogenomic Biomarkers, Centre for to maintenance of homeostatic proportions of those cell Integrative Biology, and Microsoft Research, Centre for Computational Systems types. Biology, University of Trento, Trento, Italy. Hotchkiss Brain Institute, Department of Psychiatry, University of Calgary, Calgary, AB, Canada. Departments of Psychiatry and Biobehavioral Sciences and Psychology, UCLA, Limitations Los Angeles, CA, USA. Department of Psychiatry, UCSD, San Diego, CA, USA. 8 9 Foremost among limitations, the sample sizes of the Department of Psychology, Yale University, New Haven, CT, USA. Department of Psychiatry, Zucker Hillside Hospital, Long Island, NY, USA. Department of three groups in our study were all small. We plan to add Psychiatry, UCSF and San Francisco VA Healthcare System, San Francisco, CA, data from more subjects in the near future (NAPLS3). In USA. Department of Psychiatry, Harvard Medical School at Beth Israel addition, the Luminex platform is less reproducible for Deaconess Medical Center and Massachusetts General Hospital, Boston, MA, USA. Department of Psychiatry, Center for Behavioral Genomics UCSD, San certain proteins than other platforms, including Meso- Diego, CA, USA. Departments of Psychology and Psychiatry, Emory Scale, ELISA, and microfluidic ELISA. We plan to employ University, Atlanta, GA, USA microfluidic ELISA in pending analyses. Another limita- tion is the two-year limit for the definition of conversion. Conflict of interest The authors declare that they have no conflict of interest. Although the rate of psychosis conversion is much lower after two years, we anticipate that up to 10% of our nonconverters will eventually convert to psychosis, Publisher's note Springer Nature remains neutral with regard to jurisdictional claims in implying some uncertainty in all our statements about published maps and institutional affiliations. nonconverters vs converters. However, by considering only networks constructed from extremely high correla- Supplementary Information accompanies this paper at (https://doi.org/ tions and thousands of random subsets of the groups, we 10.1038/s41398-018-0158-y). consider it unlikely that the findings could be explained Received: 23 December 2017 Revised: 6 April 2018 Accepted: 15 April 2018 entirely by chance alignments. Lastly, many proteins functionally related to those in our assays were not con- sidered due to costs. Therefore, important signals might be missing from our analyses. Despite all the seemingly References plausible relationships described supra, our findings need 1. Sabherwal, S., English, J. A., Focking, M., Cagney, G. & Cotter, D. R. Blood confirmation and expansion in other cohorts. Further biomarker discovery in drug-free schizophrenia: The contributionof pro- work of multiple types, e.g., as that being undertaken by teomics and multiplex immunoassays. Expert. Rev. Proteom. 13,1141–1155 (2016). Sorokin , is required to directly link peripheral immune 2. Miller, B. J., Buckley, P., Seabolt, W., Mellor, A. & Kirkpatrick, B. Meta-analysis of dysregulation to the etiology of schizophrenia. cytokine alterations in schizophrenia: clinical status and antipsychotic effects. Biol. Psychiatry 70,663–671 (2011). 3. Perkins, D. O. et al. Towards a psychosis risk blood diagnostic for persons Translation to the clinic experiencing high-risk symptoms: Preliminary results from the NAPLS project. The present work suggests that additional longitudinal Schizophr. Bull. 41,419–428 (2015). 4. Chan, M. K. et al. Development of a blood-based molecular biomarker test for studies of clinical high-risk patients with logically identification of schizophrenia before disease onset. Transl. Psychiatry 5,e601 expanded assays of agents of immune responses could (2015). lead to parsimonious lists of proteins implicated in con- 5. Metcalf, S. A. et al. Serum C-reactive protein in adolescence and risk of schi- zophrenia in adulthood: a prospective birth cohort study. Brain Behav. Immun. version to schizophrenia. Ratios of distinguishing proteins 59,253–259 (2017). for a new patient could be compared with historical ratios 6. Fusar-Poli, P. et al. At risk or not at risk? A meta-analysis of the prognostic to predict membership among nonconverters or accuracy of psychometric interviews for psychosis prediction. World Psychiatry 14,322–332 (2015). converters. 7. Khandaker, G. M.,Pearson, R. M., Zammit, S., Lewis,G.& Jones, P. B. Association Furthermore, the above potential connection of SER- of serum interleukin 6 and C-reactive protein in childhood with depression PINE1 and TIMP1 expression with CTGF regulation and psychosis in young adult life: a population-based longitudinal study. might lead to investigation of Pamrevlumab (http://www. JAMA Psychiatry 71, 1121–1128 (2014). Jeffries et al. Translational Psychiatry (2018) 8:112 Page 9 of 10 8. Perkins, D. O., Gu, H.,Boteva, K. &Lieberman,J.A. Relationship 35. Kirkpatrick, B. & Kennedy, B. K. Accelerated aging in schizophrenia and betweendurationofuntreated psychosis and outcome in first-episode related disorders: Future research. Schizophr. Res. in press, Available online 6 schizophrenia: a critical review and meta-analysis. Am.J.Psychiatry 162, July 2017. 1785–1804 (2005). 36. Khan, S. S. et al. A null mutation in SERPINE1 protects against biological aging 9. Domenici, E.et al. Plasma protein biomarkers for depression and schizophrenia in humans. Sci. Adv. 3, eaao1617 (2017). by multi analyte profiling of case-control collections. PLoS ONE 5,(2010).. 37. Hoirisch-Clapauch, S. & Nardi, A. E. Psychiatric remission with warfarin: should 10. Dickerson, F. et al. Multianalyte markers of schizophrenia and bipolar disorder: psychosis be addressed as plasminogen activator imbalance? Med. Hypotheses a preliminary study. Schizophr. Res. 168,450–455 (2015). 80,137–141 (2013). 11. Guest,P.C.et al. Proteomicprofiling in schizophrenia: enabling stratification 38. Erdbruegger, U., Haubitz, M. & Woywodt, A. Circulating endothelial cells: a formoreeffectivetreatment. Genome Med. 5, 25 (2013). novel marker of endothelial damage. Clin. Chim. Acta 373,17–26 (2006). 12. Ramsey, J. M. et al. Distinct molecular phenotypes in male and female schi- 39. Aird, W. C. Endothelial cell heterogeneity. Cold Spring Harb. Perspect. Med. 2, zophrenia patients. PLoS ONE 8, e78729 (2013). a006429 (2012). 13. Fredrickson,B.L.et al. Afunctional genomic perspective on human well- 40. Augustin, H. G., & Koh, G. Y. Organotypic vasculature: From descriptive het- being. Proc. Natl Acad. Sci. USA 110, 13684–13689 (2013). erogeneity to functional pathophysiology. Science 357,eaal2379(2017). 14. Cannon, T. D. et al. Prediction of psychosis in youth at high clinical risk: a 41. Shin, C. Y.,Kundel, M. & Wells, D.G.Rapid, activity-induced increase in tissue multisite longitudinal study in North America. Arch.Gen.Psychiatry 65,28–37 plasminogen activator is mediated by metabotropic glutamate receptor- (2008). dependent mRNA translation. J. Neurosci. 24,9425–9433 (2004). 15. Cannon, T. D. et al. An individualized risk calculator for research in prodromal 42. Lochner, J. E. et al. Activity-dependent release of tissue plasminogen activator psychosis. Am.J.Psychiatry 173,980–988 (2016). from the dendritic spines of hippocampal neurons revealed by live-cell ima- 16. Khandaker, G. M. et al. Inflammation and immunity in schizophrenia: impli- ging. J. Neurobiol. 66, 564–577 (2006). cations for pathophysiology and treatment. Lancet Psychiatry 2,258–270 43. Zhu, T. X. et al. ECM-related gene expression profile in vascular smooth muscle (2015). cells from human saphenous vein and internal thoracic artery. J. Cardiothorac. 17. Nave, K. A. & Ehrenreich, H. A bloody brake on myelin repair. Nature 553, Surg. 8, 155 (2013). 31–32 (2018). 44. Fudge, N. J. & Mearow, K. M. Extracellular matrix-associated gene expression in 18. Korschenhausen, D. A., Hampel, H. J., Ackenheil, M., Penning, R. & Muller, N. adult sensory neuron populations culturedonalamininsubstrate. BMC Fibrin degradation products in post mortem brain tissue of schizophrenics: a Neurosci. 14, 15 (2013). possible marker for underlying inflammatory processes. Schizophr. Res. 19, 45. Suzuki, Y.,Nagai,N.&Umemura, K. A reviewofthe mechanisms of 103–109 (1996). blood–brain barrier permeability by tissue-type plasminogen activator treat- 19. Ryu, J. K. & McLarnon, J. G. A leaky blood-brain barrier, fibrinogen infiltration ment for cerebral ischemia. Front. Cell. Neurosci. 10,2 (2016). and microglial reactivity in inflamed Alzheimer’s disease brain. J. Cell.Mol.Med. 46. Hoirisch-Clapauch, S. & Nardi, A. E. Markers of low activity of tissue plasmi- 13,2911–2925 (2009). nogen activator/plasmin are prevalent in schizophrenia patients. Schizophr. 20. Petersen, M. A. et al. Fibrinogen activates BMP signaling in oligodendrocyte Res. 159,118–123 (2014). progenitor cells and inhibits remyelination after vascular damage. Neuron 96, 47. Lemarchand, E. et al. Stressed neurons protect themselves by a tissue-type 1003–1012 (2017). e7. plasminogen activator-mediated EGFR-dependent mechanism. Cell Death 21. Weber, M. D., Godbout, J. P. & Sheridan, J. F. Repeated social defeat, neu- Differ. 23,123–131 (2016). roinflammation, and behavior: monocytes carry the signal. Neuropsycho- 48. Bertrand, T. et al. Conformations of tissue plasminogen activator (tPA) pharmacology 42,46–61 (2017). orchestrate neuronal survival by a crosstalk between EGFR and NMDAR. Cell 22. Marin,I.A.&Kipnis,J.Central nervoussystem: (Immunological) ivorytower or Death Dis. 6, e1924 (2015). not? Neuropsychopharmacology 42,28–35 (2017). 49. Deng, X. et al. Positive association of phencyclidine-responsive genes, PDE4A 23. Filiano,A.J., Gadani,S.P.&Kipnis,J.How andwhy do Tcells andtheir derived and PLAT, with schizophrenia. Am.J.Med.Genet. B. Neuropsychiatr. Genet. cytokines affect the injured and healthy brain? Nat. Rev. Neurosci. 18,375–384 156B,850–858 (2011). (2017). 50. Nicole, O. et al. The proteolytic activity of tissue-plasminogen activator 24. Addington, J. A. et al. North American Prodrome Longitudinal Study (NAPLS enhances NMDA receptor-mediated signaling. Nat. Med. 7,59–64 (2001). 2): overview and recruitment. Schizophr. Res. 142,77–82 (2012). 51. Yang, F. et al. Tissue plasminogen activator expression and barrier properties 25. Miller, T. J. et al. Prospective diagnosis of the initial prodrome for schizophrenia of human brain microvascular endothelial cells. Cell.Physiol.Biochem. 28, based on the Structured Interview for Prodromal Syndromes: preliminary 631–638 (2011). evidence of interrater reliability and predictive validity. Am. J. Psychiatry 159, 52. Reijerkerk, A. et al. Tissue-type plasminogen activator is a regulator of 863–865 (2002). monocyte diapedesis through the brain endothelial barrier. J. Immunol. 181, 26. First, M. B., Spitzer R. L., Givvon M., Williams J. B. W. Structured Clinical Interview 3567–3574 (2008). for DSM-IV TR Axis I Disorders, Non-patient Edition (SCID-I/NP) (Biometrics 53. Hoirisch-Clapauch, S. & Nardi, A. E. Improvement of psychotic symptoms Research, New York State Psychiatric Institute, New York, 2002). and the role of tissue plasminogen activator. Int. J. Mol. Sci. 16, 27550–27560 27. Johnson, N. L. Systems of frequency curves generated by methods of trans- (2015). lation. Biometrika 36,149–176 (1949). 54. Kolomeets, N. S., & Uranova, N. A. Reduced oligodendrocyte density in layer 5 28. Bonchev,D.&Buck, G.A.in Complexity in Chemistry, Biology, and Ecology (eds. of the prefrontal cortex in schizophrenia. Eur. Arch. Psychiatry Clin. Neurosci. 1–8, Bonchev, D. & Rouvray, D. H.) 191–235 (Springer US, 2005). Available online 23 March 2018. 29. Jeffries, C. D. et al. Insights into psychosis risk from leukocyte microRNA 55. Richeldi, L. Targeted treatment of idiopathic pulmonary fibrosis: one step at a expression. Transl. Psychiatry 6, e981 (2016). time. Eur. Respir. J. 47,1321–1323 (2016). 30. Najjar, S. et al. Neurovascular unit dysfunction and blood-brain barrier 56. Bonniaud, P. et al. Progressive transforming growth factor beta1-induced lung hyperpermeability contribute to schizophrenia neurobiology: a theoretical fibrosis is blocked by an orally active ALK5 kinase inhibitor. Am. J. Respir. Crit. integration of clinical and experimental evidence. Front. Psychiatry 8, 83 (2017). Care Med. 171,889–898 (2005). 31. Rijken,D. C.&Sakharov, D.V.Basic principles in thrombolysis: regulatory role of 57. Tsoutsman,T.etal. CCN2 playsakey role in extracellular matrix gene plasminogen. Thromb. Res. 103(Suppl 1), S41–S49 (2001). expression in severe hypertrophic cardiomyopathy and heart failure. J. Mol. 32. Hennekens, C. H. Increasing global burden of cardiovascular disease in general Cell. Cardiol. 62,164–178 (2013). populations and patients with schizophrenia. J. Clin. Psychiatry 68(Suppl 4), 4–7 58. Tan, J. T. et al. Connective tissue growth factor/CCN-2 is upregulated in epi- (2007). didymal and subcutaneous fat depots in a dietary-induced obesity model. Am. 33. Hippisley-Cox, J., Vinogradova, Y., Coupland, C. & Parker, C. Risk of malignancy J. Physiol. Endocrinol. Metab. 304, E1291–E1302 (2013). in patients with schizophrenia or bipolar disorder: nested case-control study. 59. Wang, X., McLennan, S. V., Allen, T. J. & Twigg, S. M. Regulation of pro- Arch. Gen. Psychiatry 64, 1368–1376 (2007). inflammatory and pro-fibrotic factors by CCN2/CTGF in H9c2 cardiomyocytes. 34. Hoirisch-Clapauch,S., Amaral,O. B., Mezzasalma,M.A., Panizzutti,R.& Nardi, A. J. Cell. Commun. Signal. 4,15–23 (2010). E. Dysfunction in the coagulation system and schizophrenia. Transl. Psychiatry 60. de Fouw, N. J. et al. The cofactor role of protein S in the acceleration of whole 6, e704 (2016). blood clot lysis by activated protein C in vitro. Blood 67,1189–1192 (1986). Jeffries et al. Translational Psychiatry (2018) 8:112 Page 10 of 10 61. Long, G. L.,Marshall, A.,Gardner,J.C.&Naylor,S.L.Genes forhuman vitamin 73. Ramasamy, R. et al. Advanced glycation end products and RAGE: a common K-dependent plasma proteins C and S are located on chromosomes 2 and 3, thread in aging, diabetes, neurodegeneration, and inflammation. Glycobiology respectively. Somat. Cell Mol. Genet. 14,93–98 (1988). 15,16R–28R (2005). 62. Rezende, S. M., Simmonds, R. E. & Lane, D. A. Coagulation, inflammation, and 74. Steullet, P. et al. Redox dysregulation, neuroinflammation, and NMDA receptor apoptosis: different roles for protein S and the protein S-C4b binding protein hypofunction: A “central hub” in schizophrenia pathophysiology? Schizophr. complex. Blood 103,1192–1201 (2004). Res. 176,41–51 (2016). 63. Agarwal,V., Talens,S., Grandits,A. M.&Blom,A. M.Anovelinteraction 75. Berretta, S.,Pantazopoulos,H., Markota, M.,Brown,C.&Batzianouli, E. T. Losing between complement inhibitor C4b-binding protein and plasminogen that the sugar coating: potential impact of perineuronal net abnormalities on enhances plasminogen activation. J. Biol. Chem. 290, 18333–18342 (2015). interneurons in schizophrenia. Schizophr. Res. 167,18–27 (2015). 64. Sekar, A. et al. Schizophrenia risk from complex variation of complement 76. Cabungcal, J. H. et al. Perineuronal nets protect fast-spiking interneurons component 4. Nature 530,177–183 (2016). against oxidative stress. Proc. Natl Acad. Sci. USA 110,9130–9135 (2013). 65. Brown,N.J. Therapeutic potentialofplasminogen activator inhibitor-1 inhi- 77. Benchenane, K. et al. Tissue-type plasminogen activator crosses the intact bitors. Ther. Adv. Cardiovasc Dis. 4, 315–324 (2010). blood-brain barrier by low-density lipoprotein receptor-related protein-medi- 66. Hebert, M., Lesept, F., Vivien, D. & Macrez, R. The story of an exceptional serine ated transcytosis. Circulation 111,2241–2249 (2005). protease, tissue-type plasminogen activator (tPA). Rev. Neurol. 172,186–197 78. Winkler, J. L., Kedees, M. H., Guz, Y. & Teitelman, G. Inhibition of connective (2016). tissue growth factor by small interfering ribonucleic acid prevents increase in 67. Hoirisch-Clapauch, S. & Nardi, A. E. Multiple roles of tissue plasminogen acti- extracellular matrix molecules in a rodent model of diabetic retinopathy. Mol. vator in schizophrenia pathophysiology. Semin. Thromb. Hemost. 39,950–954 Vis. 18,874–886 (2012). (2013). 79. Brocker, C., Thompson, D., Matsumoto, A., Nebert, D. W. & Vasiliou, V. Evolu- 68. Pang, P. T. et al. Cleavage of proBDNF by tPA/plasmin is essential for long-term tionary divergence and functions of the human interleukin (IL) gene family. hippocampal plasticity. Science 306, 487–491 (2004). Hum. Genom. 5,30–55 (2010). 69. Lin, P. Y. Regulation of proteolytic cleavage of brain-derived neurotrophic 80. Akdis, M. et al. Interleukins, from 1 to 37, and interferon-gamma: receptors, factor precursor by antidepressants in human neuroblastoma cells. Neu- functions, and roles in diseases. J. Allergy Clin. Immunol. 127,701–721 (2011). ropsychiatr. Dis. Treat. 11,2529–2532 (2015). e1-70. 70. Lepeta, K. & Kaczmarek, L. Matrix metalloproteinase-9 as a novel player in 81. Kohn, L. A. et al. Lymphoid priming in human bone marrow begins before synaptic plasticity and schizophrenia. Schizophr. Bull. 41,1003–1009 (2015). expression of CD10 with upregulation of L-selectin. Nat. Immunol. 13,963–971 71. Hashimoto, K. Sigma-1 receptor chaperone and brain-derived neurotrophic (2012). factor: emerging links between cardiovascular disease and depression. Prog. 82. Sorokin, L. The impact of the extracellular matrix on inflammation. Nat. Rev. Neurobiol. 100,15–29 (2013). Immunol. 10,712–723 (2010). 72. Turner, R. J. & Sharp, F. R. Implications of MMP9 for blood brain barrier disruption and hemorrhagic transformation following ischemic stroke. Front. Cell. Neurosci. 10, 56 (2016). http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Translational Psychiatry Springer Journals

Networks of blood proteins in the neuroimmunology of schizophrenia

Download PDF
10 pages

Loading next page...
 
/lp/springer_journal/networks-of-blood-proteins-in-the-neuroimmunology-of-schizophrenia-dMJoaklO2V

References (89)

Publisher
Springer Journals
Copyright
Copyright © 2018 by The Author(s)
Subject
Medicine & Public Health; Medicine/Public Health, general; Psychiatry; Neurosciences; Behavioral Sciences; Pharmacotherapy; Biological Psychology
eISSN
2158-3188
DOI
10.1038/s41398-018-0158-y
Publisher site
See Article on Publisher Site

Abstract

Levels of certain circulating cytokines and related immune system molecules are consistently altered in schizophrenia and related disorders. In addition to absolute analyte levels, we sought analytes in correlation networks that could be prognostic. We analyzed baseline blood plasma samples with a Luminex platform from 72 subjects meeting criteria for a psychosis clinical high-risk syndrome; 32 subjects converted to a diagnosis of psychotic disorder within two years while 40 other subjects did not. Another comparison group included 35 unaffected subjects. Assays of 141 analytes passed early quality control. We then used an unweighted co-expression network analysis to identify highly correlated modules in each group. Overall, there was a striking loss of network complexity going from unaffected subjects to nonconverters and thence to converters (applying standard, graph-theoretic metrics). Graph differences were largely driven by proteins regulating tissue remodeling (e.g. blood-brain barrier). In more detail, certain sets of antithetical proteins were highly correlated in unaffected subjects (e.g. SERPINE1 vs MMP9), as expected in homeostasis. However, for particular protein pairs this trend was reversed in converters (e.g. SERPINE1 vs TIMP1, being synthetical inhibitors of remodeling of extracellular matrix and vasculature). Thus, some correlation signals strongly predict impending conversion to a psychotic disorder and directly suggest pharmaceutical targets. Introduction endothelial cell function and general integrity. Such Circulating levels of immune system proteins and rela- findings support the hypothesis that psychosis involves ted signaling agents are consistently altered in schizo- brain dysregulation by an altered peripheral immune phrenia. This observation includes unmedicated first system and aberrant signaling at BBB. 1,2 episode psychosis patients and persons at clinical high- Persons meeting clinical high-risk criteria have about a 3–5 risk who subsequently convert to psychosis . Consistent 20% risk of developing a psychotic disorder within two findings include proteins in the immune system acute years, and that is 100-fold higher than the 0.2% general phase response and in the plasminogen activating sys- population risk . Baseline factors differentiating clinical tem . Many of the proteins influence brain function high-risk subjects that convert to psychosis vs do not directly, crossing blood–brain barrier (BBB) and signaling convert may be of etiologic significance. Previous psy- glia or other perivascular cells. These immune signaling chosis risk prediction studies compared levels of immune molecules also regulate brain function by influencing BBB signaling molecules, individually or in a linear combina- 3–5,7 tion . Since responses of the peripheral immune sys- tem are highly coordinated, investigation of the correlation patterns of immune signaling molecules might Correspondence: Clark D Jeffries ([email protected]) Renaissance Computing Institute, University of North Carolina, Chapel Hill, NC, also be informative. In this analysis we sought networks of USA highly correlated immune molecules in persons at clinical Department of Psychiatry, University of North Carolina, Chapel Hill, NC, USA high-risk who developed psychosis, compared to those Full list of author information is available at the end of the article Larry J. Seidman deceased on 7 September 2017. © The Author(s) 2018 Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to theCreativeCommons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/. 1234567890():,; 1234567890():,; 1234567890():,; 1234567890():,; Jeffries et al. Translational Psychiatry (2018) 8:112 Page 2 of 10 who did not develop psychosis over the two-year follow- A second key concept is the pleiotropic roles of proteins up period. We also analyzed data from unaffected com- customarily considered in the context of hemostasis vs parison subjects. In some respects, the correlation net- hemorrhage. Specifically, as Nave and Ehrenreich demon- works were strikingly different in converters. Some strated , “It is becoming apparent that coagulation factors distinguishing proteins were modulators of extracellular do much more than simply act in the blood-coagulation matrix (ECM) components and BBB. cascade.” For example, fibrinogen associates with schizo- 18 19 20 Schizophrenia pervasively impacts brain functions, phrenia , Alzheimer disease , and multiple sclerosis (MS) . typically causing in adolescence or early adulthood dis- A third key concept is that the peripheral immune ability that is chronic and relapsing. Improved clinical system impacts physiological and pathological brain 21–23 outcomes are often associated identification and treat- function . We have hypothesized that peripheral ment of the disorder early in its course , implying the immune system dysregulation may contribute to devel- desirability of seeking reliable predictors based upon opment of psychosis. Furthermore, the construction of readily accessible biomarkers. Moreover, early indicators the above classifier survived permutation testing very might be close to the ultimate causes of schizophrenia. well, implying that certain peripheral blood plasma pro- Thus, researchers (e.g., Domenici et al. and Dickerson teins and conversion of patients to schizophrenia are et al. ) have investigated in particular proteins and other associated. Finally, dysregulation of BBB permeability, blood plasma analytes that distinguish unaffected com- transendothelial cell migration, or remodeling of brain parison subjects from patients with schizophrenia or ECM or vasculature—all impacted by the immune system 11,12 subsets of such patients . Included among the analytes —could obviously be important in the events leading to have been concentrations of circulating cytokines and frank psychotic mood disorders including schizophrenia. other immune system signaling molecules. These can be altered in schizophrenia across all stages of the disorder Materials and methods including the prodromal stage. Subjects One key concept of this paper is information from The aims and methods of North American Prodrome networks. As Fredrickson et al. stated, “…accumulation Longitudinal Study (NAPLS2) were described in detail of many individually noisy indicator variables can yield previously . Briefly, NAPLS2 is an eight-site observa- highly stable estimates of the underlying factors they tional study of predictors and mechanisms of conversion share in common.” Thus, we sought networks of highly to psychosis. The NAPLS2 cohort includes 765 subjects at correlated signals among each of three groups (converters, clinical high-risk for psychosis, based on the Criteria of nonconverters, and unaffected comparison subjects) from Prodromal States, determined by the Structured Interview our North American Prodrome Longitudinal Study for Prodromal Syndromes, and rated with the Scale of 14 25 (NAPLS) project . The emphasis was analysis of data Prodromal Symptoms (see Supplement). In addition, the collected at the prodrome state (see Supplement Figure cohort includes 280 demographically similar unaffected S1). To place the present work in context, NAPLS is a comparison subjects. Subjects were between ages 12 and multi-site program that has accumulated, cleaned, stored, 35 at baseline. Psychiatric diagnoses were determined by analyzed, and reported many types of clinical and the Structured Clinical Interview for DSM IV . Clinical laboratory assays, leading to proposals of various pre- assessments were performed every six months, and sub- dictors and mechanisms for the development of jects were followed for up to two years. The study was psychosis . approved by the Institutional Review Board at each site, Thus, the present emphasis differs from some earlier and each subject provided written informed consent or works in that it is longitudinal and it pertains entirely to assent, with a parent or guardian also consenting for networks. The three networks from the three groups are minor subjects. Demographics for subjects in the present formed by proteins that are highly correlated over subjects study are provided in Table 1. in each group, far more highly correlated than could be reasonably explained by chance. All of the reported cor- Plasma analytes relations are positive because there we observed no During a baseline visit, blood samples were drawn using negative correlations of the same, very high magnitude. Becton Dickenson P100 blood collection tubes containing Many of the distinguishing analytes we found turned EDTA (as anticoagulant), proprietary protein stabilizers, out to be prominent in the immune system and its and a mechanical separator. Mean processing time to interaction with trophic factors and ECM components. −80 °C storage was 28 min (SD = 2 min). Plasma samples Thus, the proper arena of this paper became neu- were subsequently sent on dry ice to Myriad RBM (Austin roimmunology. Particularly in the last decade, many TX), a laboratory that has maintained CLIA Accreditation important reports have developed this view of mental since 2006. Samples were analyzed with the Human illnesses (e.g., Khandaker et al. ). DiscoveryMAP v. 1.0 assays of 185 analytes associated Jeffries et al. Translational Psychiatry (2018) 8:112 Page 3 of 10 Table 1 Demographic and clinical characteristics of study comparison subjects. Quality control, normalization subjects taken at baseline of longitudinal study methods, and tests with duplicated samples were per- formed, as previously described . Unaffected Clinical high-risk, Clinical high- comparison nonconverters risk, Data analyses (UC) N = 35 (CHR-NC) N = 40 Converters We used an unweighted co-expression network analysis (CHR-C) N = to identify highly correlated networks of analytes in each of the three groups. Pearson correlations were calculated using macros and built-in functions in Excel. Limited by Age, average (SD) 20 (4.5) 19.5 (4.6) 19.2 (3.7) the smallest group size (converters, n = 32), we performed Ancestry random re-sampling with replacement 10,000 times by %Caucasian 60%, 65% 55% drawing subgroups of 28 subjects from each of the three %African 31% 17.5% 21% groups. In other words, the 9870 pairs of analytes were compared in 28-dimensional space 10,000 times to cal- %Asian 9% 17.5% 24% culate Pearson correlations. We compared those correla- Sex, % female 34% 37.5% 30.3% tion values to a common threshold and then recorded the SES, average (SD) 4.8 (1.8) 4.5 (2.3) 4.5 (1.8) number of times a pair exceeded the threshold (flowcharts Time blood draw, 12.:12 pm 12:39 pm (2.0 h) 11:59 am in Supplement Figure S2). The correlation threshold average (SD) (1.85 h) (1.79 h) selected was 0.7662; if two 28-dimensional vectors were populated by a Gaussian distribution, then the probability Prescription medication of their correlation exceeding this value would be 1E-6. Antipsychotic 0% 25% 13% Thus, with 9870 pairs of analytes, we expect only 1 such Antidepressant 1% 30% 25% value among all pairs about once in 100 trials. Stimulant 0% 8% 6% Considering all 107 subjects and all 9870 pairs of ana- lytes at once, distribution analysis using EasyFit Mood stabilizer 0% 5% 3% (MathWave, Dnepropetrovsk, Ukraine) revealed that no Benzodiazepine 0% 5% 13% normal distribution fit the observations but the Johnson NSAID 0% 0% 0% S distribution did. Technically, normal distributions Antibiotic 0% 0% 0% failed to fit the observed distribution of correlations in both Kolmogorov–Smirnov and Anderson–Darling tests Substance Use for even α = 0.01; by contrast, the Johnson S distribution Tobacco use 9% 30% 44% achieves α = 0.2 in both tests. The Johnson S distribu- Alcohol use 46% 48% 38% tion accommodates a one-sided tail of very strong, posi- Marijuana use 9% 25% 31% tive correlations not balanced by even one negative correlation of equal magnitude. Therefore, we focused on Current co-morbid DSM IV Diagnosis only the very strong, positive correlations (Fig. 1). Depression 0% 45% 50% We validated our approach by first randomly permuting Anxiety 3% 60% 56% all 107 subjects, then analyzing 28-subject subsets of the g,* Disorders pseudo-groups exactly as if they were from the true data. Neither the design nor the implementation of the algo- CHR-C vs UC FET p-value = 0.047, CHR-NC vs UC Fisher Exact Test (FET) p-value = 0.001 rithm conferred any obvious, consistent distinction to CHR-C vs UC FET p-value = 0.011, CHR-NC vs UC FET p-value = 0.002 edge counts of graphs of very strong correlations in the CHR-C vs UC FET p-value = 0.047 CHR-C vs UC FET p-value = 0.001, CHR-NC vs UC FET p-value = 0.02 three pseudo-groups. This result indicates that both the CHR-C vs UC FET p-value = 0.020, CHR-NC vs UC FET p-value = 0.056 design and implementation of the program correctly CHR-C vs UC FET p-value < 0.0001, CHR-NC vs UC FET p-value < 0.0001 CHR-C vs UC FET p-value < 0.0001, CHR-NC vs UC FET p-value < 0.0001 achieved unbiased identification of correlation networks. *Depression disorders include Major Depression, Depressive Disorder Not Measures of network complexity include those of Otherwise Specified, and Dysthymic Disorder. Anxiety Disorders include Obsessive Compulsive Disorder, Post-Traumatic Stress Disorder, Panic Disorder, Bonchev and Buck (see their formulas 19a,b) in which Agoraphobia, Social Phobia, Specific Phobia, Generalized Anxiety Disorder the normalized edge complexity is the number of (undirected) edges e divided by the maximum possible with immune system function, hormonal responses, oxi- number of edges using the same number of vertices v, dative stress, and metabolism. In total 141 analytes passed namely, v*(v-1)/2. preliminary quality control and were used in our analyses. Protein network analyses were conducted using Inge- We standardized to z-scores the results for each analyte nuity Pathway Analysis (IPA) (QIAGEN N.V., Venlo, The using the average and SD values of the unaffected Netherlands). Jeffries et al. Translational Psychiatry (2018) 8:112 Page 4 of 10 Results draws of 28 subjects from each group are shown for Protein correlation networks distinguish the three groups unaffected subjects (Fig. 2), nonconverters (Fig. 3), and There was a striking loss of network complexity going converters (Fig. 4). from unaffected to nonconverters and thence to converter For unaffected, nonconverter, and converter groups we subjects. Analyte pairs with Pearson correlations exceed- calculated a simple metric of graph complexity, namely ing an a priori threshold in at least 5000 of 10,000 random the ratio of numbers of edges to vertices; respective values were 1.48, 1.30, and 0.89 edges per vertex. Normalized edge complexities were 0.134, 0.119, and 0.062. Moreover, it is visually apparent that the three graphs are quite dif- ferent in terms of edge densities. Additional graphs and experiments with alternative choices of thresholds led to the same trend and are in the Supplement. This trend toward simplicity going from unaffected to nonconverter to converter for plasma proteins parallels the same for leukocytic microRNAs reported earlier . Notably, in unaffected subjects there was a network involving the proteins SERPINE1 (plasminogen activator inhibitor-1), MMP7, MMP9 (activated), MMP9t (total), and MMP10 (matrix metalloproteinases). While SER- PINE1 is anti-remodeling, the others are pro-remodeling, and a balance (correlation of some pairs) might be expected in hemostasis vs hemorrhage. In contrast, non- converters and converters lacked these strong SERPINE1 correlations. Furthermore, while the important inhibitor of remodeling TIMP1 (tissue inhibitor of metalloprotei- nases) was absent (at the chosen threshold for inclusion) Fig. 1 A histogram with 20 bins of 9870 correlation values from strong correlations in unaffected and nonconverter among 141 analytes over all 107 subjects. The shown distribution is a Johnson S fit with four parameters: gamma= −1.2967, delta= graphs, it generally exceeded the threshold correlation 2.2624, lambda= 0.26593, xi= −0.12371. The present study is value with SERPINE1 in converter subjects. In detail, the distinguished from many others by focusing on the tail of very strong, SERPINE1 vs TIMP1 correlations exceeded the threshold positive correlations (blue box) that are not balanced by any negative in the three sets of 10,000 random draws of 28 subjects in correlations of the same magnitude totals of <10, <10, and >8300 times, respectively. As sets, Fig. 2 Unaffected comparison subject data yielded a graph of strongly correlated analytes with 23 analytes (vertices) and 34 robust correlations (edges). Blood plasma proteins are labeled by their gene common symbols. The correlations in orange appear in all three graphs. We note eight analytes that include SERPINE1, and in particular SERPINE1 correlations include the matrix metalloproteinases MMP7, MMP9, and MMP10. MMP9t denotes an assay for both pro-MMP9 and mature MMP9 Jeffries et al. Translational Psychiatry (2018) 8:112 Page 5 of 10 Fig. 3 Nonconverter data yielded a graph of strongly correlated analytes with 23 analytes and 30 robust correlations. SERPINE1 correlations are completely absent, suggesting a loss of requlation of expression of the gene Fig. 4 Converter data yielded a graph of strongly correlated analytes with 27 analytes and 24 robust correlations. Four SERPINE1 correlations are present, suggesting changes in requlation of the SERPINE1 gene compared to unaffected and nonconverter assays. Remarkable is the gained correlation of SERPINE1 and TIMP1 because both proteins inhibit anticogulation and vascular remodeling in some contexts and both generally promote anti-inflammation. Furthermore, TIMP1 is completely absent in unaffected and nonconverter graphs. Also, the strong SERPINE1 correlations with matrix metalloproteinases (MMPs) in the unaffected graph are absent in this converter graph the actual SERPINE1 and TIMP1 correlation values over is known to promote mutual expression of SERPINE1 and 10,000 draws for converter subjects (mean correlation = TIMP1 and so might be part of the explanation (see 0.79, SD = 0.03) tended to be significantly greater (p- Discussion). value < 1E-100) than the values in nonconverters (mean From a previous study by Domenici el al. we evaluated = 0.55, SD = 0.06) and unaffected subjects (mean = 0.50, SERPINE1 vs TIMP1 correlations. The subjects con- SD = 0.09). The dramatic distinction of SERPINE1 and tributing to that study were schizophrenia patients and TIMP1 correlations in converters is further illustrated in controls. We calculated correlations for 10,000 random the Supplement (Figure S10). subsets of 200 subjects drawn randomly from 267 control What could explain the extremely high correlations of subjects and 229 schizophrenia patients. The results SERPINE1 and TIMP1 in converters only? We note that (Supplement Figure S11) indicated patients had far another protein, CTGF (connective tissue growth factor), stronger SERPINE1 vs TIMP1 correlations than controls. Jeffries et al. Translational Psychiatry (2018) 8:112 Page 6 of 10 Subnetworks common among all three groups psychosis cannot be simple. However, Hoirisch-Clapauch 34,46 Several pairs of analytes were highly correlated in all et al. . provided additional reasoning and evidence of a three groups (highlighted in orange in Figs. 2, 3, and 4). connection, finding lower PLAT levels among 70 schizo- Among them was the complete graph formed by inter- phrenia patients compared with 98 age-matched controls. leukins 3, 5, 7, and 13. We examined the correlation Since the principal regulator of PLAT levels is SERPINE1, 9 46 pattern of these interleukins in the Domenici data for 267 Hoirisch-Clapauch and Nardi postulated that inflam- unaffected subjects. As shown in the Supplement Figure matory conditions could increase the risk of schizo- S9, the four interleukins were highly correlated in that phrenia through mechanisms involving SERPINE1 levels. cohort as well. PLAT is highly pleiotropic as, therefore, might be some of its regulators. Two of many reported PLAT substrates Discussion directly relevant to psychiatry have functions in epidermal Immune correlation networks and psychosis risk growth factor receptor (EGFR)-mediated neuroprotec- 47,48 Psychosis conversion in high-risk subjects was char- tion and the dynamics of the glutamate receptor and acterized by marked simplifications of networks of cor- ion channel protein NMDAR (antagonized by phencycli- 49,50 related proteins that regulate tissue remodeling; this dine in a model of schizophrenia) . Furthermore, might be consistent with the hypothesis of BBB dysre- PLAT is expressed in brain endothelial cells, neurons, and gulation in schizophrenia . Notable were graph changes microglia. It is strongly and negatively correlated with involving SERPINE1. The serine protease inhibitor SER- mRNA for OCLN (occludin) , a protein important in the PINE1 is the major inhibitor of the plasminogen activator stability and permeability of tight junctions. Reactive PLAT; PLAT activates plasminogen to plasmin (the final oxygen species can cause rapid release of PLAT in substrate of the fibrinolytic system) which in turn monocyte and endothelia cocultures . In a related effect degrades fibrin. Thus, a focus infra is SERPINE1 and (noting the above correlations in unaffected subjects), related proteins . MMP9 promotes monocyte migration through brain Relevant to the present study, schizophrenia patients are parenchymal basement membrane . subject to increased risk of cardiovascular disease, Consistent with our findings, there is also an emerging reduced risk of certain types of cancer, and possibly some literature implicating the plasminogen pathway in schi- aspects of accelerated aging; all these general observations zophrenia. This literature points towards down-regulation have in turn been associated with SERPINE1 or related of PLAT, up-regulation of factors including SERPINE1 32–36 34,37,46,53 proteins that inhibit PLAT activation . , and consequences More particularly, connection of schizophrenia with of elevated fibrinogen (fibrin) including incorrect differ- proteins important in coagulation homeostasis has been entiation of progenitors of oligodendrocytes into astro- amplified by recent studies of Hoirisch-Clapauch, Nardi, cytes in MS . This misguided differentiation might relate and their colleagues , starting with anecdotal descrip- to reported low levels of oligodendrocytes in layer five of tions of possible benefits to five schizophrenia patients of prefrontal cortex in schizophrenia . treatment with warfarin (which would seemingly counter For data from the same subjects presently studied we the effects of increased SERPINE1 and decreased PLAT previously reported that proteins involved in the plasmi- activity). They proposed that normalization of levels of nogen pathway, including MMP7 and Factor 7, predicted PLAT in the psychotic brain might enable long-term psychosis conversion . In our previous publication we did remission of psychotic symptoms. not find that expression levels of SERPINE1 or TIMP1 PLAT is secreted by many cell types including blood were substantially increased going from nonconverters to vessel endothelial cells that are known to express diverse converters. By contrast, the current study suggests that genes differentially even among regions of the same the extreme changes in correlation networks of SER- 38–40 artery . In neurons PLAT occurs in dendrites and PINE1 and related proteins, especially TIMP1, may synapses; it is stored in pre-synaptic vesicles that cross the indeed be informative. synaptic cleft after depolarization, subject to regulation by Most notable is an interpretation of CTGF in idiopathic 41,42 55,56 57–59 astrocytes . pulmonary fibrosis and other diseases that might Aside from hemostasis, PLAT has related roles in explain the remarkable correlation of SERPINE1 and 31,43 smooth muscle signaling in blood vessel tunica , sev- TIMP1 levels in converters. Unfortunately, in our blood eral types of ECM signaling , and BBB permeability in plasma assays CTGF was not detected in 50–70% of the the context of cerebral ischemia . Given the large num- subjects in all three of the groups, precluding its corre- bers of schizophrenia patients who are comorbid for lation analysis. However, regarding CTGF as a known cardiovascular diseases, who require treatment with driver of SERPINE1 and TIMP1 levels, we note that local warfarin or functionally related pharmaceuticals, and who upregulation of CTGF by a type of alveolar cell, resulting are not in remission, the intersection of hemostasis and in proliferation of local fibroblasts, has been proposed as a Jeffries et al. Translational Psychiatry (2018) 8:112 Page 7 of 10 surrogate biomarker of idiopathic pulmonary fibrosis MMP9 is also an important factor in ECM maintenance (IPF); CTGF is therefore the target of ongoing IPF and remodeling, assayed herein both as mature MMP9 clinical trials of the fully human monoclonal antibody and as the total of proMMP9 and mature MMP9 (labeled Pamrevlumab by FibroGen (San Francisco CA) (trial MMP9t in our graphs). It regulates glutamate receptors, NCT01890265 at ClinicalTrials.gov). modulates physiological and morphological synaptic The serine protease inhibitor SERPINE1 is a “suicide plasticity, and is regulated by glutamate at excitatory synapses protein” in the serpin superfamily and is the major inhi- . MMP9 and other extracellular proteases such bitor of the plasminogen activator PLAT and related as plasmin convert proBDNF to mature BDNF . PLAT, protein PLAU (plasminogen activator urokinase). Thus, PLAU, plasmin, and MMP9 all have roles in BBB dis- SERPINE1 is a crucial down regulator of fibrinolysis and ruption after stroke . MMP9 itself is activated in pre- ECM degradation . In wound healing, SERPINE1 con- sence of oxidative stress and in turn promotes the tributes to regulation of cell proliferation and tissue receptor for advanced glycation end-product (RAGE) (to remodeling , as well as other aspects of cell signaling and be described in a forthcoming paper by D. Dwir et al.), 44,45 migration . said to induce a self-reinforcing cycle of inflammatory 73,74 Studies of samples from clinical high-risk psychiatric responses and further oxidative stress . In addition, patients in the literature and in our data (Supplement MMP9 is involved in the degradation of the perineuronal Figure S10) can exhibit somewhat elevated levels of net, a type of ECM that wraps fast-spiking parvalbumin SERPINE1. Interestingly, de Fouw et al. reported that interneurons; the perineuronal net is known to be affected the glycoprotein Protein S (PROS1) accelerated neu- in the medial prefrontal brain of schizophrenia 75,76 tralization of SERPINE1, thereby increasing PLAT activ- patients . In mice, intravenously administered PLAT ity; this finding suggests possibilities for indirect control was detected within the brain parenchyma and cere- of PLAT levels and function. In its free form, PROS1 is a brospinal fluid, having crossed the BBB by transcytosis . cofactor of protein C (PROC) in the anticoagulation In summary, for converter subjects SERPINE1 was no pathway. PROS1 and PROC are vitamin K-dependent longer correlated with antithetical MMPs, but instead plasma proteins , included among the cofactors in the became highly correlated with the synthetical TIMP1. inactivation of the prothrombinase complex. PROS1 also This observation and the above integration of these pro- exists in a complex with complement C4B-binding pro- teins and CTGF in multiple disorders suggest the poten- tein . Complement inhibitor C4B-binding protein tial importance of CTGF inhibition. Several inhibitors of CTGF have been proposed and studied; for example, enhances plasminogen activation . Isotypes of the C4 protein have been implicated in schizophrenia gene insulin and an siRNA have been studied in the context of association studies . inhibition of vascular remodeling in diabetic retino- SERPINE1 levels are up-regulated by inflammatory pathy . However, certain advantages attend a mono- cytokines including IL6, TNF, and TGFB1, as reviewed by clonal antibody for CTGF inhibition. Brown et al. . These researchers postulated that inflam- matory conditions increased the risk of schizophrenia Interleukin networks common in all subjects through a combination of mechanisms involving We found that IL3, IL5, IL7, and IL13 form a highly increased SERPINE1 levels or decreased free PROS1 correlated module in all groups, and replicated this finding levels. in an external test set of unaffected subjects (Figure S8), PLAT was not measured in our assays, but more should suggesting co-regulation and closely related functions of 66,67 be said about some of its extensive functions . PLAT these cytokines. Three genes, IL3, IL5, and IL13 (and also degrades ECM and in particular cleaves the large glyco- IL4), exist as a cytokine gene cluster on chromosome arm protein RELN, all contributing to cell migration and tissue 5q. Unfortunately, IL4 detection was weak (below LLOQ) remodeling, particularly revascularization. PLAT affects or missing in more than half our samples, precluding neural, endothelial, and glial cells, driving multiple, correlation analysis with the rest of that cluster. But IL4 sometimes opposing effects by activating or otherwise and IL7 are among the six “γ-chain utilizing” interleukins, regulating diverse transcription factors and receptors. while IL3, IL4, IL5, and IL13 are the four “IL4-like” Regarding the neurotrophin BDNF (brain-derived neu- interleukins; both sets of proteins are considered tightly rotrophic factor), PLAT promotes the cleavage of neu- packed α-helices in a four-helix bundle motif of short core ronal proBDNF to mature BDNF, a function essential for helices . There are about 45 human interleukins, so the late-phase long-term potentiation . Pro- and mature members of the identified correlation cluster are more BDNF have important and opposing effects on synaptic similar than would be expected by chance. plasticity, regulation of neurogenesis, and neuronal sur- The top shared pathway for IL3, IL5, IL7, and IL13 from vival . Only the mature BDNF was assayed by us; it is application of IPA was “hematopoiesis from multipotent prominent in Figs. 2–4. stem cells” (p = 1.6E-9). IL3 promotes hematopoietic Jeffries et al. Translational Psychiatry (2018) 8:112 Page 8 of 10 stem cell proliferation, and IL7 regulates development of fibrogen.com/pamrevlumab-trials/) or related agents of the common lymphoid progenitor cells and development CTGF inhibition. The potential medical value of inhibi- of B-cell, T-cell, and NK-cell lineages. In detail, IL3 reg- tion of CTGF in various diseases has long been recog- ulates development of the common myeloid progenitor, nized. The recent appearance of Pamrevlumab as a and IL5 regulates development of neutrophils and eosi- potential treatment for IPF and its repurposing suggested nophils . Less is known regarding IL13, but its receptor, herein for prevention of conversion to psychosis might be especially important. Repurposing is a drug development IL12A1, is selectively expressed on CD62L+ cells (com- mon lymphoid progenitors), again suggesting a role in strategy with huge savings of time and expense. regulation of lymphoid cell differentiation . Normally, Author details the circulating proportions of neutrophils, lymphocytes, Renaissance Computing Institute, University of North Carolina, Chapel Hill, NC, monocytes, and eosinophils are relatively constant, so USA. Department of Psychiatry, University of North Carolina, Chapel Hill, NC, USA. Department of Psychiatry, Lausanne University Hospital (CHUV), possibly IL3, IL5, IL7, and IL13 in our graphs contribute Lausanne, Switzerland. Laboratory of Neurogenomic Biomarkers, Centre for to maintenance of homeostatic proportions of those cell Integrative Biology, and Microsoft Research, Centre for Computational Systems types. Biology, University of Trento, Trento, Italy. Hotchkiss Brain Institute, Department of Psychiatry, University of Calgary, Calgary, AB, Canada. Departments of Psychiatry and Biobehavioral Sciences and Psychology, UCLA, Limitations Los Angeles, CA, USA. Department of Psychiatry, UCSD, San Diego, CA, USA. 8 9 Foremost among limitations, the sample sizes of the Department of Psychology, Yale University, New Haven, CT, USA. Department of Psychiatry, Zucker Hillside Hospital, Long Island, NY, USA. Department of three groups in our study were all small. We plan to add Psychiatry, UCSF and San Francisco VA Healthcare System, San Francisco, CA, data from more subjects in the near future (NAPLS3). In USA. Department of Psychiatry, Harvard Medical School at Beth Israel addition, the Luminex platform is less reproducible for Deaconess Medical Center and Massachusetts General Hospital, Boston, MA, USA. Department of Psychiatry, Center for Behavioral Genomics UCSD, San certain proteins than other platforms, including Meso- Diego, CA, USA. Departments of Psychology and Psychiatry, Emory Scale, ELISA, and microfluidic ELISA. We plan to employ University, Atlanta, GA, USA microfluidic ELISA in pending analyses. Another limita- tion is the two-year limit for the definition of conversion. Conflict of interest The authors declare that they have no conflict of interest. Although the rate of psychosis conversion is much lower after two years, we anticipate that up to 10% of our nonconverters will eventually convert to psychosis, Publisher's note Springer Nature remains neutral with regard to jurisdictional claims in implying some uncertainty in all our statements about published maps and institutional affiliations. nonconverters vs converters. However, by considering only networks constructed from extremely high correla- Supplementary Information accompanies this paper at (https://doi.org/ tions and thousands of random subsets of the groups, we 10.1038/s41398-018-0158-y). consider it unlikely that the findings could be explained Received: 23 December 2017 Revised: 6 April 2018 Accepted: 15 April 2018 entirely by chance alignments. Lastly, many proteins functionally related to those in our assays were not con- sidered due to costs. Therefore, important signals might be missing from our analyses. Despite all the seemingly References plausible relationships described supra, our findings need 1. Sabherwal, S., English, J. A., Focking, M., Cagney, G. & Cotter, D. R. Blood confirmation and expansion in other cohorts. Further biomarker discovery in drug-free schizophrenia: The contributionof pro- work of multiple types, e.g., as that being undertaken by teomics and multiplex immunoassays. Expert. Rev. Proteom. 13,1141–1155 (2016). Sorokin , is required to directly link peripheral immune 2. Miller, B. J., Buckley, P., Seabolt, W., Mellor, A. & Kirkpatrick, B. Meta-analysis of dysregulation to the etiology of schizophrenia. cytokine alterations in schizophrenia: clinical status and antipsychotic effects. Biol. Psychiatry 70,663–671 (2011). 3. Perkins, D. O. et al. Towards a psychosis risk blood diagnostic for persons Translation to the clinic experiencing high-risk symptoms: Preliminary results from the NAPLS project. The present work suggests that additional longitudinal Schizophr. Bull. 41,419–428 (2015). 4. Chan, M. K. et al. Development of a blood-based molecular biomarker test for studies of clinical high-risk patients with logically identification of schizophrenia before disease onset. Transl. Psychiatry 5,e601 expanded assays of agents of immune responses could (2015). lead to parsimonious lists of proteins implicated in con- 5. Metcalf, S. A. et al. Serum C-reactive protein in adolescence and risk of schi- zophrenia in adulthood: a prospective birth cohort study. Brain Behav. Immun. version to schizophrenia. Ratios of distinguishing proteins 59,253–259 (2017). for a new patient could be compared with historical ratios 6. Fusar-Poli, P. et al. At risk or not at risk? A meta-analysis of the prognostic to predict membership among nonconverters or accuracy of psychometric interviews for psychosis prediction. World Psychiatry 14,322–332 (2015). converters. 7. Khandaker, G. M.,Pearson, R. M., Zammit, S., Lewis,G.& Jones, P. B. Association Furthermore, the above potential connection of SER- of serum interleukin 6 and C-reactive protein in childhood with depression PINE1 and TIMP1 expression with CTGF regulation and psychosis in young adult life: a population-based longitudinal study. might lead to investigation of Pamrevlumab (http://www. JAMA Psychiatry 71, 1121–1128 (2014). Jeffries et al. Translational Psychiatry (2018) 8:112 Page 9 of 10 8. Perkins, D. O., Gu, H.,Boteva, K. &Lieberman,J.A. Relationship 35. Kirkpatrick, B. & Kennedy, B. K. Accelerated aging in schizophrenia and betweendurationofuntreated psychosis and outcome in first-episode related disorders: Future research. Schizophr. Res. in press, Available online 6 schizophrenia: a critical review and meta-analysis. Am.J.Psychiatry 162, July 2017. 1785–1804 (2005). 36. Khan, S. S. et al. A null mutation in SERPINE1 protects against biological aging 9. Domenici, E.et al. Plasma protein biomarkers for depression and schizophrenia in humans. Sci. Adv. 3, eaao1617 (2017). by multi analyte profiling of case-control collections. PLoS ONE 5,(2010).. 37. Hoirisch-Clapauch, S. & Nardi, A. E. Psychiatric remission with warfarin: should 10. Dickerson, F. et al. Multianalyte markers of schizophrenia and bipolar disorder: psychosis be addressed as plasminogen activator imbalance? Med. Hypotheses a preliminary study. Schizophr. Res. 168,450–455 (2015). 80,137–141 (2013). 11. Guest,P.C.et al. Proteomicprofiling in schizophrenia: enabling stratification 38. Erdbruegger, U., Haubitz, M. & Woywodt, A. Circulating endothelial cells: a formoreeffectivetreatment. Genome Med. 5, 25 (2013). novel marker of endothelial damage. Clin. Chim. Acta 373,17–26 (2006). 12. Ramsey, J. M. et al. Distinct molecular phenotypes in male and female schi- 39. Aird, W. C. Endothelial cell heterogeneity. Cold Spring Harb. Perspect. Med. 2, zophrenia patients. PLoS ONE 8, e78729 (2013). a006429 (2012). 13. Fredrickson,B.L.et al. Afunctional genomic perspective on human well- 40. Augustin, H. G., & Koh, G. Y. Organotypic vasculature: From descriptive het- being. Proc. Natl Acad. Sci. USA 110, 13684–13689 (2013). erogeneity to functional pathophysiology. Science 357,eaal2379(2017). 14. Cannon, T. D. et al. Prediction of psychosis in youth at high clinical risk: a 41. Shin, C. Y.,Kundel, M. & Wells, D.G.Rapid, activity-induced increase in tissue multisite longitudinal study in North America. Arch.Gen.Psychiatry 65,28–37 plasminogen activator is mediated by metabotropic glutamate receptor- (2008). dependent mRNA translation. J. Neurosci. 24,9425–9433 (2004). 15. Cannon, T. D. et al. An individualized risk calculator for research in prodromal 42. Lochner, J. E. et al. Activity-dependent release of tissue plasminogen activator psychosis. Am.J.Psychiatry 173,980–988 (2016). from the dendritic spines of hippocampal neurons revealed by live-cell ima- 16. Khandaker, G. M. et al. Inflammation and immunity in schizophrenia: impli- ging. J. Neurobiol. 66, 564–577 (2006). cations for pathophysiology and treatment. Lancet Psychiatry 2,258–270 43. Zhu, T. X. et al. ECM-related gene expression profile in vascular smooth muscle (2015). cells from human saphenous vein and internal thoracic artery. J. Cardiothorac. 17. Nave, K. A. & Ehrenreich, H. A bloody brake on myelin repair. Nature 553, Surg. 8, 155 (2013). 31–32 (2018). 44. Fudge, N. J. & Mearow, K. M. Extracellular matrix-associated gene expression in 18. Korschenhausen, D. A., Hampel, H. J., Ackenheil, M., Penning, R. & Muller, N. adult sensory neuron populations culturedonalamininsubstrate. BMC Fibrin degradation products in post mortem brain tissue of schizophrenics: a Neurosci. 14, 15 (2013). possible marker for underlying inflammatory processes. Schizophr. Res. 19, 45. Suzuki, Y.,Nagai,N.&Umemura, K. A reviewofthe mechanisms of 103–109 (1996). blood–brain barrier permeability by tissue-type plasminogen activator treat- 19. Ryu, J. K. & McLarnon, J. G. A leaky blood-brain barrier, fibrinogen infiltration ment for cerebral ischemia. Front. Cell. Neurosci. 10,2 (2016). and microglial reactivity in inflamed Alzheimer’s disease brain. J. Cell.Mol.Med. 46. Hoirisch-Clapauch, S. & Nardi, A. E. Markers of low activity of tissue plasmi- 13,2911–2925 (2009). nogen activator/plasmin are prevalent in schizophrenia patients. Schizophr. 20. Petersen, M. A. et al. Fibrinogen activates BMP signaling in oligodendrocyte Res. 159,118–123 (2014). progenitor cells and inhibits remyelination after vascular damage. Neuron 96, 47. Lemarchand, E. et al. Stressed neurons protect themselves by a tissue-type 1003–1012 (2017). e7. plasminogen activator-mediated EGFR-dependent mechanism. Cell Death 21. Weber, M. D., Godbout, J. P. & Sheridan, J. F. Repeated social defeat, neu- Differ. 23,123–131 (2016). roinflammation, and behavior: monocytes carry the signal. Neuropsycho- 48. Bertrand, T. et al. Conformations of tissue plasminogen activator (tPA) pharmacology 42,46–61 (2017). orchestrate neuronal survival by a crosstalk between EGFR and NMDAR. Cell 22. Marin,I.A.&Kipnis,J.Central nervoussystem: (Immunological) ivorytower or Death Dis. 6, e1924 (2015). not? Neuropsychopharmacology 42,28–35 (2017). 49. Deng, X. et al. Positive association of phencyclidine-responsive genes, PDE4A 23. Filiano,A.J., Gadani,S.P.&Kipnis,J.How andwhy do Tcells andtheir derived and PLAT, with schizophrenia. Am.J.Med.Genet. B. Neuropsychiatr. Genet. cytokines affect the injured and healthy brain? Nat. Rev. Neurosci. 18,375–384 156B,850–858 (2011). (2017). 50. Nicole, O. et al. The proteolytic activity of tissue-plasminogen activator 24. Addington, J. A. et al. North American Prodrome Longitudinal Study (NAPLS enhances NMDA receptor-mediated signaling. Nat. Med. 7,59–64 (2001). 2): overview and recruitment. Schizophr. Res. 142,77–82 (2012). 51. Yang, F. et al. Tissue plasminogen activator expression and barrier properties 25. Miller, T. J. et al. Prospective diagnosis of the initial prodrome for schizophrenia of human brain microvascular endothelial cells. Cell.Physiol.Biochem. 28, based on the Structured Interview for Prodromal Syndromes: preliminary 631–638 (2011). evidence of interrater reliability and predictive validity. Am. J. Psychiatry 159, 52. Reijerkerk, A. et al. Tissue-type plasminogen activator is a regulator of 863–865 (2002). monocyte diapedesis through the brain endothelial barrier. J. Immunol. 181, 26. First, M. B., Spitzer R. L., Givvon M., Williams J. B. W. Structured Clinical Interview 3567–3574 (2008). for DSM-IV TR Axis I Disorders, Non-patient Edition (SCID-I/NP) (Biometrics 53. Hoirisch-Clapauch, S. & Nardi, A. E. Improvement of psychotic symptoms Research, New York State Psychiatric Institute, New York, 2002). and the role of tissue plasminogen activator. Int. J. Mol. Sci. 16, 27550–27560 27. Johnson, N. L. Systems of frequency curves generated by methods of trans- (2015). lation. Biometrika 36,149–176 (1949). 54. Kolomeets, N. S., & Uranova, N. A. Reduced oligodendrocyte density in layer 5 28. Bonchev,D.&Buck, G.A.in Complexity in Chemistry, Biology, and Ecology (eds. of the prefrontal cortex in schizophrenia. Eur. Arch. Psychiatry Clin. Neurosci. 1–8, Bonchev, D. & Rouvray, D. H.) 191–235 (Springer US, 2005). Available online 23 March 2018. 29. Jeffries, C. D. et al. Insights into psychosis risk from leukocyte microRNA 55. Richeldi, L. Targeted treatment of idiopathic pulmonary fibrosis: one step at a expression. Transl. Psychiatry 6, e981 (2016). time. Eur. Respir. J. 47,1321–1323 (2016). 30. Najjar, S. et al. Neurovascular unit dysfunction and blood-brain barrier 56. Bonniaud, P. et al. Progressive transforming growth factor beta1-induced lung hyperpermeability contribute to schizophrenia neurobiology: a theoretical fibrosis is blocked by an orally active ALK5 kinase inhibitor. Am. J. Respir. Crit. integration of clinical and experimental evidence. Front. Psychiatry 8, 83 (2017). Care Med. 171,889–898 (2005). 31. Rijken,D. C.&Sakharov, D.V.Basic principles in thrombolysis: regulatory role of 57. Tsoutsman,T.etal. CCN2 playsakey role in extracellular matrix gene plasminogen. Thromb. Res. 103(Suppl 1), S41–S49 (2001). expression in severe hypertrophic cardiomyopathy and heart failure. J. Mol. 32. Hennekens, C. H. Increasing global burden of cardiovascular disease in general Cell. Cardiol. 62,164–178 (2013). populations and patients with schizophrenia. J. Clin. Psychiatry 68(Suppl 4), 4–7 58. Tan, J. T. et al. Connective tissue growth factor/CCN-2 is upregulated in epi- (2007). didymal and subcutaneous fat depots in a dietary-induced obesity model. Am. 33. Hippisley-Cox, J., Vinogradova, Y., Coupland, C. & Parker, C. Risk of malignancy J. Physiol. Endocrinol. Metab. 304, E1291–E1302 (2013). in patients with schizophrenia or bipolar disorder: nested case-control study. 59. Wang, X., McLennan, S. V., Allen, T. J. & Twigg, S. M. Regulation of pro- Arch. Gen. Psychiatry 64, 1368–1376 (2007). inflammatory and pro-fibrotic factors by CCN2/CTGF in H9c2 cardiomyocytes. 34. Hoirisch-Clapauch,S., Amaral,O. B., Mezzasalma,M.A., Panizzutti,R.& Nardi, A. J. Cell. Commun. Signal. 4,15–23 (2010). E. Dysfunction in the coagulation system and schizophrenia. Transl. Psychiatry 60. de Fouw, N. J. et al. The cofactor role of protein S in the acceleration of whole 6, e704 (2016). blood clot lysis by activated protein C in vitro. Blood 67,1189–1192 (1986). Jeffries et al. Translational Psychiatry (2018) 8:112 Page 10 of 10 61. Long, G. L.,Marshall, A.,Gardner,J.C.&Naylor,S.L.Genes forhuman vitamin 73. Ramasamy, R. et al. Advanced glycation end products and RAGE: a common K-dependent plasma proteins C and S are located on chromosomes 2 and 3, thread in aging, diabetes, neurodegeneration, and inflammation. Glycobiology respectively. Somat. Cell Mol. Genet. 14,93–98 (1988). 15,16R–28R (2005). 62. Rezende, S. M., Simmonds, R. E. & Lane, D. A. Coagulation, inflammation, and 74. Steullet, P. et al. Redox dysregulation, neuroinflammation, and NMDA receptor apoptosis: different roles for protein S and the protein S-C4b binding protein hypofunction: A “central hub” in schizophrenia pathophysiology? Schizophr. complex. Blood 103,1192–1201 (2004). Res. 176,41–51 (2016). 63. Agarwal,V., Talens,S., Grandits,A. M.&Blom,A. M.Anovelinteraction 75. Berretta, S.,Pantazopoulos,H., Markota, M.,Brown,C.&Batzianouli, E. T. Losing between complement inhibitor C4b-binding protein and plasminogen that the sugar coating: potential impact of perineuronal net abnormalities on enhances plasminogen activation. J. Biol. Chem. 290, 18333–18342 (2015). interneurons in schizophrenia. Schizophr. Res. 167,18–27 (2015). 64. Sekar, A. et al. Schizophrenia risk from complex variation of complement 76. Cabungcal, J. H. et al. Perineuronal nets protect fast-spiking interneurons component 4. Nature 530,177–183 (2016). against oxidative stress. Proc. Natl Acad. Sci. USA 110,9130–9135 (2013). 65. Brown,N.J. Therapeutic potentialofplasminogen activator inhibitor-1 inhi- 77. Benchenane, K. et al. Tissue-type plasminogen activator crosses the intact bitors. Ther. Adv. Cardiovasc Dis. 4, 315–324 (2010). blood-brain barrier by low-density lipoprotein receptor-related protein-medi- 66. Hebert, M., Lesept, F., Vivien, D. & Macrez, R. The story of an exceptional serine ated transcytosis. Circulation 111,2241–2249 (2005). protease, tissue-type plasminogen activator (tPA). Rev. Neurol. 172,186–197 78. Winkler, J. L., Kedees, M. H., Guz, Y. & Teitelman, G. Inhibition of connective (2016). tissue growth factor by small interfering ribonucleic acid prevents increase in 67. Hoirisch-Clapauch, S. & Nardi, A. E. Multiple roles of tissue plasminogen acti- extracellular matrix molecules in a rodent model of diabetic retinopathy. Mol. vator in schizophrenia pathophysiology. Semin. Thromb. Hemost. 39,950–954 Vis. 18,874–886 (2012). (2013). 79. Brocker, C., Thompson, D., Matsumoto, A., Nebert, D. W. & Vasiliou, V. Evolu- 68. Pang, P. T. et al. Cleavage of proBDNF by tPA/plasmin is essential for long-term tionary divergence and functions of the human interleukin (IL) gene family. hippocampal plasticity. Science 306, 487–491 (2004). Hum. Genom. 5,30–55 (2010). 69. Lin, P. Y. Regulation of proteolytic cleavage of brain-derived neurotrophic 80. Akdis, M. et al. Interleukins, from 1 to 37, and interferon-gamma: receptors, factor precursor by antidepressants in human neuroblastoma cells. Neu- functions, and roles in diseases. J. Allergy Clin. Immunol. 127,701–721 (2011). ropsychiatr. Dis. Treat. 11,2529–2532 (2015). e1-70. 70. Lepeta, K. & Kaczmarek, L. Matrix metalloproteinase-9 as a novel player in 81. Kohn, L. A. et al. Lymphoid priming in human bone marrow begins before synaptic plasticity and schizophrenia. Schizophr. Bull. 41,1003–1009 (2015). expression of CD10 with upregulation of L-selectin. Nat. Immunol. 13,963–971 71. Hashimoto, K. Sigma-1 receptor chaperone and brain-derived neurotrophic (2012). factor: emerging links between cardiovascular disease and depression. Prog. 82. Sorokin, L. The impact of the extracellular matrix on inflammation. Nat. Rev. Neurobiol. 100,15–29 (2013). Immunol. 10,712–723 (2010). 72. Turner, R. J. & Sharp, F. R. Implications of MMP9 for blood brain barrier disruption and hemorrhagic transformation following ischemic stroke. Front. Cell. Neurosci. 10, 56 (2016).

Journal

Translational Psychiatry – Springer Journals

Published: Jun 6, 2018

There are no references for this article.