A high-throughput proteo-genomics method to identify antibody
targets associated with malignant disease
Yinyin Huang, Joseph Franklin, Kathreen Gifford, Bruce L. Roberts,
and Charles A. Nicolette
Applied Genomics, Genzyme Corporation, Framingham, MA 01701-9322, USA
Received 19 December 2003; accepted 23 December 2003
Identification of antibody targets associated with malignant disease is indispensable to developing passive and active antibody-based
therapeutics or diagnostic agents. We have developed a novel technique combining Western blotting, genetic profiling, and mass
spectroscopy that allows for the rapid and unambiguous identification of such antigens in a high-throughput manner. Herein, we demonstrate
this technique, designated Ab SCAN, by deducing the known target of a monoclonal antibody and by identifying a new antigen that was
observed to be the frequent target of humoral immune responses in prostate cancer patients. In both instances, a specific antigen emerged as
the sole protein candidate. The newly identified antigen, mannose-6-phosphate/IGF II receptor, may be an important naturally immunogenic
antigen involved in prostate cancer. The Ab SCAN technique is uniquely suited to the analysis of longitudinal serum samples from clinical
studies and could be a powerful tool to correlate humoral immune responses directed against discreet antigens with clinical events.
D 2004 Elsevier Inc. All rights reserved.
Keywords: spectroscopy; Ab SCAN; antibody
The identification of tumor-associated antibody targets
can enable new therapeutic, diagnostic, and prognostic
options for the diagnosis and management of malignant
disease. Following the development of humanized mono-
clonal antibody technologies, treatments based on geneti-
cally engineered monospecific antibodies directed at tumor-
associated proteins have been clinically validated with the
success of herceptin  and rituxan . Consequently,
several other promising antibody-based therapies are cur-
rently under development [3 –6]. We believe that additional
therapeutic targets can be identified by characterizing nat-
ural or induced humoral immune responses involved in
malignant disease. Dividends from this approach will hinge
on the ability to identify the targeted antigens, particularly
those that correlate with favorable clinical outcomes.
Many laboratories are involved in the effort to charac-
terize antibody reactivities present in cancer patient serum.
Techniques such as SEREX  and protein microarrays 
have been successful at identifying these targets, however,
these techniques can be laborious, time-consuming, and
expensive. Furthermore, these methods lack the flexibility
to analyze longitudinal serum samples in a high-throughput
manner. We have termed our methodology Antibody-based
Systematic Characterization of Antigens (Ab SCAN). Ab
SCAN was developed to be both high-throughput and
economical in terms of serum consumption.
A schematic of the Ab SCAN methodology is shown in
Fig. 1. The first step involves Western blotting patient serum
against a panel of human tumor cell line lysates to visualize
reactivities of interest and identify their respective apparent
molecular weights. Next, a band of interest is excised from a
gel identical to that used to create the Western blot. After
digesting the proteins within the gel slice with trypsin,
MALDI-TOF mass spectroscopic analysis measures the
masses of the tryptic peptides which are used to reconstruct
the proteins present in the gel slice. The cell lines used on
the Western blot have been genetically profiled via SAGE
analysis  thus providing a quantitative measure of all
genes expressed by these cells in culture. By examining the
SAGE-predicted expression pattern for each candidate pro-
tein on the MALDI-TOF MS list, the correct antigen can be
readily identified because it will be statistically likely that
1521-6616/$ - see front matter D 2004 Elsevier Inc. All rights reserved.
* Corresponding author. Applied Genomics, Genzyme Corporation, 1
Mountain Road, Framingham, MA 01701-9322. Fax: +1-508-620-1203.
E-mail address: Bruce.Roberts@genzyme.com (B.L. Roberts).
Clinical Immunology 111 (2004) 202 – 209