Characterization of monoclonal antibodies raised against the latent nuclear antigen of human herpesvirus 8

Abstract

Human herpesvirus 8 (HHV-8; also designated Kaposi's sarcoma-associated herpesvirus) is the likely etiological agent of Kaposi's sarcoma (KS). HHV-8 encodes a latent nuclear antigen (LNA) which is the product of the viral gene orf 73. LNA is recognized by most infected patient sera and is the basis of current immunofluorescence assays used in epidemiological studies of HHV-8 infection. Here we describe the characterization of four monoclonal antibodies raised to the C-terminal third of LNA-glutathione S-transferase fusion proteins. These monoclonal antibodies recognized discrete linear epitopes within the C terminus and repetitive region of LNA, detected antigen in primary effusion lymphoma (PEL) cells, and precipitated a 220- to 230-kDa protein doublet corresponding to LNA from HHV-8-infected PEL cell lines. In situ immunocytochemistry of KS lesions with these antibodies show that LNA is extensively expressed in KS spindle cells.

FIG. 1

Fusion protein mapping of LNA MAbs. The C-terminal locations of six GST-LNA fusion proteins (GST-C14, -C7, -C17, -C11, -C9, and -C10) and two histidine-tagged fusion proteins (TH-C14 and TH-NN) are shown relative to that of full-length LNA. Numbers represent amino acids relative to the putative start methionine at codon 127296 of the published HHV-8 genome (26). The hatched box represents the extent of the repetitive coding region of LNA. Restriction enzyme sites NruI, AgeI, and StuI used to construct the fusion protein clones are indicated. The reactivity of each MAb to each fusion protein is shown as A₄₁₀ above background of greater than 0.5 absorbance units (+++), between 0.1 and 0.5 absorbance units (++), or between 0.02 and 0.1 absorbance units (+). The epitope-mapped position of each MAb is indicated on TH-C14. The epitopes are EQEQE for LN53, EVDYPV for LN72, and THPKKPHPRYQQ for LN69.

FIG. 2

Immunoprecipitation and Western blot analyses using LNA MAbs. (a) Immunoprecipitates from 10⁷ BC-3 cells or Ramos cells per lane were resolved on SDS–8% polyacrylamide gels. LN53 was able to precipitate a 220- to 230-kDa doublet corresponding to LNA. Pooled antibody isotype controls to HIV gp120 for each LNA MAb as well as Protein G-Sepharose-only immunoprecipitation controls (Control and Protein G lanes, respectively) were used in identical immunoprecipitations. (b) Western blot of 2 × 10⁵ cell equivalents of total protein extract from BC-3 (2) and Ramos cells detected with LNA MAb LN20. A doublet of 220 to 230 kDa, and a smaller reactive band of approximately 180 kDa (marked on blot) are seen. Identical lanes were detected with an anti-rat alkaline phosphatase-conjugated secondary antibody as a control. The same results were obtained for LN53 and LN72 (data not shown).

FIG. 3

LNA MAbs react with nuclear bodies in PEL cells. MAbs LN20 (a), LN53 (b), and LN72 (c) all react by IFA to nuclear antigenic structures in BCP-1 cells (3, 14) identical to those detected with HHV-8-positive KS patient serum (f). Antibody LN53 did not react with nuclear antigens present in control Daudi (d) or Ramos (e) cells. LN20 and LN72 were also negative on control cells (not shown). The bar in panel a applies to all panels.

FIG. 4

FACS profile of LNA MAbs on HHV-8-infected cells. The MAbs were tested by FACS on cell lines BCP-1, BC-3, HBL-6, Daudi, and Ramos; 5 × 10⁵ cells were permeabilized, and LNA antibodies were used at the dilution of 1:100. The cells were incubated for 1 h with individual antibodies in PBS–0.05% saponin–1% FCS), washed three times in PBS–0.1% (vol/vol) Tween, and incubated for 1 h with rabbit anti-rat fluorescein isothiocyanate-conjugated (DAKO) secondary antibody. Analysis was performed on a FACS 440 (Becton Dickinson). Fluorescent intensity is expressed on an arbitrary logarithmic scale. Black histograms represent antibody-stained cells, and white histograms represent secondary antibody conjugate controls. (a) All antibodies detected the LNA-expressing cell population in BCP-1 cells. (b) Antibody LN53 does not detect LNA in the negative control cells Ramos (EBV negative) and Daudi (EBV positive) but does detect the LNA-positive population in the HHV-8-infected BC-3 and HBL-6 cell lines. All other antibodies are negative for the control cell lines (data not shown).

FIG. 5

LN53 reactivity to LNA in nodular KS. Immunocytochemistry was performed on a paraffin-embedded tumor from classical KS. Permeabilization was performed by microwave. Negative control biopsies were from angiosarcomas and hemangiomas (not shown). After incubation with normal rabbit serum (DAKO), MAbs were applied for 1 h at 22°C followed by two washes in PBS–0.1% (vol/vol) Tween. The secondary biotin-conjugated antibody (rabbit anti-rat; DAKO) was applied for 30 min followed by washing. Antibody reactions were visualized with streptavidine-alkaline phosphatase (Vector Laboratories) and a substrate red chromogen (Vector). Adjacent sections were stained by standard methods. (a) Hematoxylin and eosin staining showing KS nodule (left) and surrounding dermis (right). (b) Adjacent sections stained with LN53 showing clear demarcation between LNA expression in the KS lesion and not in the surrounding dermis (magnification, ×40). Almost all spindle-shaped cells are positive for LNA (×60) (c), with a stippling pattern reminiscent of the staining of PEL cells (×160) (d).