Mapping PAM4 (clivatuzumab), a monoclonal antibody in clinical trials for early detection and therapy of pancreatic ductal adenocarcinoma, to MUC5AC mucin

Abstract

Background: PAM4, an antibody that has high specificity for pancreatic ductal adenocarcinoma (PDAC), compared to normal pancreas, benign lesions of the pancreas, and cancers originating from other tissues, is being investigated as a biomarker for early detection, as well as antibody-targeted imaging and therapy. Therefore, the identity of the antigen bound by this monoclonal antibody (MAb) can provide information leading to improved use of the antibody. Prior results suggested the antigen is a mucin-type glycoprotein rich in cysteine disulfide bridges that provide stable conformation for the PAM4-epitope.

Methods: Indirect and sandwich enzyme immunoassays (EIA) were performed to compare and contrast the reactivity of PAM4 with several anti-mucin antibodies having known reactivity to specific mucin species (e.g., MUC1, MUC4, MUC5AC, etc.). Studies designed to block reactivity of PAM4 with its specific antigen also were performed.

Results: We demonstrate that MAbs 2-11 M1 and 45 M1, each reactive with MUC5AC, are able to provide signal in a heterologous sandwich immunoassay where PAM4 is the capture antibody. Further, we identify MAbs 21 M1, 62 M1, and 463 M1, each reactive with MUC5AC, as inhibiting the reaction of PAM4 with its specific epitope. MAbs directed to MUC1, MUC3, MUC4, MUC16 and CEACAM6 are not reactive with PAM4-captured antigen, nor are they able to block the reaction of PAM4 with its antigen.

Conclusions: These data implicate MUC5AC as a specific mucin species to which PAM4 is reactive. Furthermore, this realization may allow for the improvement of the current PAM4 serum-based immunoassay for detection of early-stage PDAC by the application of anti-MUC5AC MAbs as probes in this sandwich EIA.

Figure 1

Reactivity of several anti-mucin MAbs with a high molecular weight mucin containing fraction (CPM1) isolated from the Capan-1 human pancreatic adenocarcinoma. MAbs are identified by clone name with reactive species of mucin indicated by horizontal bars beneath MAb clone names (MUC1, etc.). In addition to PAM4, substantial reactions were observed for anti-MUC1, -MUC5AC, and -CEACAM6 antibodies. All MAbs were employed at a concentration of 10 μg/mL.

Figure 2

Reaction of several anti-mucin MAbs with PAM4-captured antigen. Mucin antigens were captured on hPAM4 coated plates, and then probed with several murine anti-mucin MAbs for reaction signal. Both anti-MUC5AC MAbs (2-11 M1 and 45 M1) bound to the hPAM4-captured mucin, whereas the anti-MUC1 MAbs (MA5 and KC4) did not bind. The homologous hPAM4/mPAM4, capture/probe immunoassay gave no signal, suggesting the density of PAM4 epitopes within the mucin may be low, possibly only a single site. A rabbit polyclonal anti-CPM1 IgG was used as a positive control for reaction with hPAM4-captured antigen.

Figure 3

Inhibition of hPAM4/antigen binding reaction by murine anti-mucin MAbs. A) Anti-mucin mMAbs (purified IgG) were added to CPM1-coated plates as potential inhibitors prior to addition of hPAM4. mPAM4 provided almost complete inhibition of the reaction between hPAM4 and antigen with the 45 M1 anti-MUC5AC providing limited inhibitory effect (IC_max = 25.5%). Neither 2-11 M1, anti-MUC5AC nor MA5 and KC4, anti-MUC1 MAbs were able to inhibit the specific hPAM4/antigen reaction. B) A similar inhibition study was performed with several anti-MUC5AC MAbs obtained as ascites fluids. MAbs 21 M1, 62 M1, and 463 M1, anti-MUC5AC provided substantial inhibitory effect, similar to that observed with mPAM4 IgG self-inhibition. The ascites form of 45 M1 yielded an inhibitory effect similar to that of the purified IgG. Ascites containing anti-alpha fetoprotein was employed as a negative control.

Figure 4

Representation of the domains of the MUC5AC glycoprotein with reactive epitopes indicated for several anti-MUC5AC MAbs. (Derived from references [22,23,28]). Data derived by transfection with plasmid vectors containing the cDNA of the 3’-end of MUC5AC, along with derivative cDNA vectors obtained by restriction enzyme digestion, have identified the location of specific epitopes for anti-MUC5AC MAbs employed in the current studies. Specific blocking studies (Figures 3A and B) suggest the PAM4-epitope resides within the cysteine-rich C-terminus domain.