Tissue-specific restriction of cyclophilin A-independent HIV-1- and SIV-derived lentiviral vectors.

The host factor alpha isoform of the tripartite motif 5 (TRIM5alpha) restricts human immunodeficiency virus type 1 (HIV-1) infection in certain non-human primate species. Restriction of HIV-1 is enhanced by binding of the viral capsid to cyclophilin A (CypA) in target cells, although CypA is not absolutely required for restriction in rhesus macaque cells. Simian immunodeficiency virus (SIV) is not restricted by rhesus macaque TRIM5alpha and its capsid does not bind to CypA. Here, the effect of lentiviral CypA dependence on restriction in different tissues was examined by engineering an HIV-1 capsid quadruple mutant (V(86)P/H(87)Q/I(91)V/M(96)I) lentiviral vector (HIV(quad)) that is CypA-independent. Whereas HIV-1 was restricted in rhesus macaque and owl monkey epithelial cells, infection with the HIV(quad) vector was efficient at high viral concentrations. In contrast, HIV(quad) was largely restricted in primary rhesus macaque CD34(+) cells. Human epithelial and primary CD34(+) cells were permissive for HIV-1, HIV(quad) and SIV, whereas transduction of human T cells by HIV(quad) or SIV was impaired. The restrictive human cells did not express increased levels of TRIM5alpha, and restriction was not relieved by abolishing CypA, consistent with HIV(quad) and SIV being CypA-independent. Pseudotyping of lentiviral vectors with the gibbon ape leukemia virus envelope altered their sensitivity to perturbations of the virus-CypA interaction compared to pseudotyping with vesicular stomatitis virus glycoproteins, suggesting that the viral entry pathway modulates restriction. Together, these studies reveal that an HIV-1 capsid quadruple mutant can partially overcome lentiviral restriction in non-human primate epithelial cells, but not in hematopoietic cells. Similarly, human cells vary in their permissiveness for CypA-independent lentiviruses, and suggest the presence of tissue-specific factor(s) that can inhibit lentiviral transduction independently of viral interaction with TRIM5alpha and CypA.

Vesicular stomatitis virus G-pseudotyped lentivirus vectors mediate efficient apical transduction of polarized quiescent primary alveolar epithelial cells.

We investigated the use of lentivirus vectors for gene transfer to quiescent alveolar epithelial cells. Primary rat alveolar epithelial cells (AEC) grown on plastic or as polarized monolayers on tissue culture-treated polycarbonate semipermeable supports were transduced with a replication-defective human immunodeficiency virus-based lentivirus vector pseudotyped with the vesicular stomatitis virus G (VSV-G) protein and encoding an enhanced green fluorescent protein reporter gene. Transduction efficiency, evaluated by confocal microscopy and quantified by fluorescence-activated cell sorting, was dependent on the dose of vector, ranging from 4% at a multiplicity of infection (MOI) of 0.1 to 99% at an MOI of 50 for AEC grown on plastic. At a comparable titer and MOI, transduction of these cells by a similarly pseudotyped murine leukemia virus vector was approximately 30-fold less than by the lentivirus vector. Importantly, comparison of lentivirus-mediated gene transfer from the apical or basolateral surface of confluent AEC monolayers (R(t) > 2 kOmega. cm(2); MOI = 10) revealed efficient transduction only when VSV-G-pseudotyped lentivirus was applied apically. Furthermore, treatment with EGTA to increase access to the basolateral surface did not increase transduction of apically applied virus, indicating that transduction was primarily via the apical membrane domain. In contrast, differentiated tracheal epithelial cells were transduced by apically applied lentivirus only in the presence of EGTA and at a much lower overall efficiency (approximately 15-fold) than was observed for AEC. Efficient transduction of AEC from the apical cell surface supports the feasibility of using VSV-G-pseudotyped lentivirus vectors for gene transfer to the alveolar epithelium and suggests that differences exist between upper and lower airways in the polarity of available receptors for the VSV-G protein.

Genomic stability of murine leukemia viruses containing insertions at the Env-3' untranslated region boundary.

Retroviruses containing inserts of exogenous sequences frequently eliminate the inserted sequences upon spread in susceptible cells. We have constructed replication-competent murine leukemia virus (MLV) vectors containing internal ribosome entry site (IRES)-transgene cassettes at the env-3' untranslated region boundary in order to examine the effects of insert sequence and size on the loss of inserts during viral replication. A virus containing an insertion of 1.6 kb replicated with greatly attenuated kinetics relative to wild-type virus and lost the inserted sequences in a single infection cycle. In contrast, MLVs containing inserts of 1.15 to 1.30 kb replicated with kinetics only slightly attenuated compared to wild-type MLV and exhibited much greater stability, maintaining their genomic integrity over multiple serial infection cycles. Eventually, multiple species of deletion mutants were detected simultaneously in later infection cycles; once detected, these variants rapidly dominated the population and thereafter appeared to be maintained at a relative equilibrium. Sequence analysis of these variants identified preferred sites of recombination in the parental viruses, including both short direct repeats and inverted repeats. One instance of insert deletion through recombination with an endogenous retrovirus was also observed. When specific sequences involved in these recombination events were eliminated, deletion variants still arose with the same kinetics upon virus passage and by apparently similar mechanisms, although at different locations in the vectors. Our results suggest that while lengthened, insert-containing genomes can be maintained over multiple replication cycles, preferential deletions resulting in loss of the inserted sequences confer a strong selective advantage.

Sequences in the cytoplasmic tail of the gibbon ape leukemia virus envelope protein that prevent its incorporation into lentivirus vectors.

Pseudotyping retrovirus and lentivirus vectors with different viral fusion proteins is a useful strategy to alter the host range of the vectors. Although lentivirus vectors are efficiently pseudotyped by Env proteins from several different subtypes of murine leukemia virus (MuLV), the related protein from gibbon ape leukemia virus (GaLV) does not form functional pseudotypes. We have determined that this arises because of an inability of GaLV Env to be incorporated into lentivirus vector particles. By exploiting the homology between the GaLV and MuLV Env proteins, we have mapped the determinants of incompatibility in the GaLV Env. Three modifications that allowed GaLV Env to pseudotype human immunodeficiency virus type 1 particles were identified: removal of the R peptide (C-terminal half of the cytoplasmic domain), replacement of the whole cytoplasmic tail with the corresponding MuLV region, and mutation of two residues upstream of the R peptide cleavage site. In addition, we have previously proposed that removal of the R peptide from MuLV Env proteins enhances their fusogenicity by transmitting a conformational change to the ectodomain of the protein (Y. Zhao et al., J. Virol. 72:5392-5398, 1998). Our analysis of chimeric MuLV/GaLV Env proteins provides further evidence in support of this model and suggests that proper Env function involves both interactions within the cytoplasmic tail and more long-range interactions between the cytoplasmic tail, the membrane-spanning region, and the ectodomain of the protein.

Receptor-specific targeting mediated by the coexpression of a targeted murine leukemia virus envelope protein and a binding-defective influenza hemagglutinin protein.

The entry of retroviral vectors into cells requires two events: binding to a cell surface receptor and the subsequent fusion of viral and cellular membranes. The host range of a vector is therefore determined largely by the receptor specificity of the fusion protein contained in the outer viral envelope. Previous attempts to generate targeted retroviral vectors have included the addition of targeting ligands to the murine leukemia virus envelope protein (MuLV Env). Although such proteins frequently display modified cell-binding characteristics, the interaction with the targeted receptors fails to trigger virus-cell fusion. Here, we report the use of a binding-defective but fusion-competent hemagglutinin (HA) protein to complement the fusion defect in a chimeric MuLV Env targeted to the Flt-3 receptor. Retroviral vectors containing both proteins showed enhanced transduction of cells expressing Flt-3, which was abrogated by preincubating the target cells with soluble Flt-3 ligand. Furthermore, the fusion function of HA was absolutely required. These data demonstrate that it is possible to separate the binding and fusion events of retroviral entry, using two separate proteins, and suggest that varying the binding protein component in this scheme may allow a general strategy for targeting retroviral vectors.

Identification of regions in the Moloney murine leukemia virus SU protein that tolerate the insertion of an integrin-binding peptide.

Targeting of retroviral vectors to specific cells has been attempted through engineering of the surface (SU) protein of the murine leukemia viruses (MuLVs), but in many cases this has adversely affected protein function and targeted delivery has been difficult to achieve. In this study, we have inserted a 15-mer peptide that binds specifically to the alpha(v)beta(3) integrin into the Moloney MuLV SU protein, including regions that are surface exposed in the crystal structure of the ecotropic receptor-binding domain. We have concentrated in particular on the variable regions VRA, VRB, and VRC, which are responsible for the use of distinct cellular receptors by different MuLV subtypes and therefore may be more likely to accommodate a heterologous binding moiety. Despite these considerations, only 8 of 26 insertion sites were tolerated, including two separate regions in VRA, a cluster of sites in VRC, and previously identified sites at the N-terminus of the protein and in the proline-rich region immediately downstream of the receptor-binding domain. When expressed on retroviral vector particles, all of the viable proteins retained the ability to bind to and transduce murine cells, although the VRC mutants and an insertion in VRA gave reduced binding and titer. Finally, although all of the viable chimeras could bind to alpha(v)beta(3) in a solid-phase binding assay, we were unable to demonstrate expanded tropism for alpha(v)beta(3)-expressing human cells. This study highlights the difficulty of engineering the Moloney MuLV SU protein, even when structural information is available, and provides guidelines for the insertion of peptide ligands into the SU protein.

Targeting retroviral vectors to CD34-expressing cells: binding to CD34 does not catalyze virus-cell fusion.

We have attempted to engineer murine leukemia virus (MuLV)-based retroviral vectors to specifically transduce cells expressing human CD34, an antigen present on the surface of undifferentiated hematopoietic stem cells. A number of chimeric ecotropic MuLV envelope (Env) proteins were constructed that contained anti-CD34 single-chain antibody variable fragments (scFvs). The scFv-Env proteins were generated either by replacing the receptor-binding domain of Env with the scFv or by inserting the scFv into the N terminus of the Env protein. Only chimeric Env proteins with scFv insertions between amino acids 6 and 7 were incorporated into viral particles, and coexpression of native MuLV Env did not rescue incorporation-defective proteins. In addition, the efficiency of incorporation varied with the specific anti-CD34 scFv that was used. Retroviral vectors containing the scFv-Env proteins bound to CD34+ cells and transduced NIH 3T3 cells expressing human CD34 (3T3-CD34 cells) at approximately twice the efficiency of the parental NIH 3T3 cells. However, the introduction of the mutation D84K, which prevents binding to the ecotropic MuLV receptor mcat-1, prevented transduction of both NIH 3T3 and 3T3-CD34 cells. Complementation cell-cell fusion assays [Zhao et al. (1997). J. Virol. 71, 6967-6972] in 3T3-CD34 cells revealed that although the scFv-Env proteins could contribute postbinding entry functions when bound to mcat-1, they were unable to do so when bound to CD34. Taken together, these data suggest that although the interaction with CD34 effectively increased the concentration of virus on 3T3-CD34 cells, entry could occur only through an interaction with mcat-1; CD34 alone was not capable of triggering the appropriate postbinding changes that lead to viral entry.

Role of variable regions A and B in receptor binding domain of amphotropic murine leukemia virus envelope protein.

For the amphotropic murine leukemia virus (MuLV), a 208-amino-acid amino-terminal fragment of the surface unit (SU) of the envelope glycoprotein is sufficient to bind to its receptor, Pit2. Within this binding domain, two hypervariable regions, VRA and VRB, have been proposed to be important for receptor recognition. In order to specifically locate residues that are important for the interaction with Pit2, we generated a number of site-specific mutations in both VRA and VRB and analyzed the resulting envelope proteins when expressed on retroviral vectors. Concurrently, we substituted portions of the amphotropic SU with homologous regions from the polytropic MuLV envelope protein. The amphotropic SU was unaffected by most of the point mutations we introduced. In addition, the deletion of eight residues in a region of VRA that was previously suggested to be essential for Pit2 utilization only decreased titer on NIH 3T3 cells by 1 order of magnitude. Although the replacement of the amino-terminal two-thirds of VRA with the polytropic sequence abolished receptor binding, smaller nonoverlapping substitutions did not affect the function of the protein. We were not able to identify a single critical receptor contact point within VRA, and we suggest that the amphotropic receptor binding domain probably makes multiple contacts with the receptor and that the loss of some of these contacts can be tolerated.

Characterization of the proline-rich region of murine leukemia virus envelope protein.

Mammalian type C retroviral envelope proteins contain a variable proline-rich region (PRR), located between the N-terminal receptor-binding domain and the more highly conserved C-terminal portion of the surface (SU) subunit. We have investigated the role of the PRR in the function of murine leukemia virus (MuLV) envelope protein. In the MuLVs, the PRR contains a highly conserved N-terminal sequence and a hypervariable C-terminal sequence. Despite this variability, the amphotropic PRR could functionally substitute for the ecotropic PRR. The hypervariable region of the PRR was not absolutely required for envelope protein function. However, truncations in this region resulted in decreased levels of both the SU and TM proteins in viral particles and increased amounts of the uncleaved precursor protein, Pr85. In contrast, the N-terminal conserved region was essential for viral infectivity. Deletion of this region prevented the stable incorporation of envelope proteins into viral particles in spite of normal envelope protein processing, wild-type levels of cell surface expression, and a wild-type ability to induce syncytia in an XC cell cocultivation assay. However, higher levels of the SU protein were shed into the supernatant, suggesting a defect in SU-TM interactions. Our data are most consistent with a role for the PRR in stabilizing the overall structure of the protein, thereby affecting the proper processing of Pr85, SU-TM interactions, and the stable incorporation of envelope proteins into viral particles. In addition, we have demonstrated that the PRR can tolerate the insertion of a peptide-binding domain, making this a potentially useful site for constructing targetable retroviral vectors.

Functional domains in the retroviral transmembrane protein.

The envelope glycoproteins of the mammalian type C retroviruses consist of two subunits, a surface (SU) protein and a transmembrane (TM) protein. SU binds to the viral receptor and is thought to trigger conformational changes in the associated TM protein that ultimately lead to the fusion of viral and host cell membranes. For Moloney murine leukemia virus (MoMuLV), the envelope protein probably exists as a trimer. We have previously demonstrated that the coexpression of envelope proteins that are individually defective in either the SU or TM subunits can lead to functional complementation (Y. Zhao et al., J. Virol. 71:6967-6972, 1997). We have now extended these studies to investigate the abilities of a panel of fusion-defective TM mutants to complement each other. This analysis identified distinct complementation groups within TM, with implications for interactions between different regions of TM in the fusion process. In viral particles, the C-terminal 16 amino acids of the MoMuLV TM (the R peptide) are cleaved by the viral protease, resulting in an increased fusogenicity of the envelope protein. We have examined the consequences of R peptide cleavage for the different TM fusion mutants and have found that this enhancement of fusogenicity can only occur in cis to certain of the TM mutants. These results suggest that R peptide cleavage enhances the fusogenicity of the envelope protein by influencing the interaction of two distinct regions in the TM ectodomain.

Mutational analysis of the fusion peptide of Moloney murine leukemia virus transmembrane protein p15E.

Fusion peptides are hydrophobic sequences located at the N terminus of the transmembrane (TM) envelope proteins of the orthomyxoviruses and paramyxoviruses and several retroviruses. The Moloney murine leukemia virus TM envelope protein, p15E, contains a hydrophobic stretch of amino acids at its N terminus followed by a region rich in glycine and threonine residues. A series of single amino acid substitutions were introduced into this region, and the resulting proteins were examined for their abilities to be properly processed and transported to the cell surface and to induce syncytia in cells expressing the ecotropic receptor. One substitution in the hydrophobic core and several substitutions in the glycine/threonine-rich region that prevented both cell-cell fusion and the transduction of NIH 3T3 cells when incorporated into retroviral vector particles were identified. In addition, one mutation that enhanced the fusogenicity of the resulting envelope protein was identified. The fusion-defective mutants trans dominantly interfered with the ability of the wild-type envelope protein to cause syncytium formation in a cell-cell fusion assay, although no trans-dominant inhibition of transduction was observed. Certain substitutions in the hydrophobic core that prevented envelope protein processing were also found. These data indicate that the N-terminal region of p15E is important both for viral fusion and for the correct processing and cell surface expression of the viral envelope protein.

Identification of envelope protein residues required for the expanded host range of 10A1 murine leukemia virus.

The 10A1 murine leukemia virus (MuLV) is a recombinant type C retrovirus isolated from a mouse infected with amphotropic MuLV (A-MuLV). 10A1 and A-MuLV have 91% amino acid identity in their envelope proteins yet display different host ranges. For example, CHO-K1 cells are resistant to A-MuLV but susceptible to infection by 10A1. We have now determined that retroviral vectors bearing altered A-MuLV envelope proteins containing 10A1-derived residues at positions 71 (A71G), 74 (Q74K), and 139 (V139M) transduce CHO-K1 cells at efficiencies similar to those achieved with 10A1 enveloped vectors. A-MuLV enveloped retroviral vectors with these three 10A1 residues were also able to transduce A-MuLV-infected NIH 3T3 cells. This observation is consistent with the ability of vectors bearing this altered A-MuLV envelope protein to recognize the 10A1-specific receptor present on NIH 3T3 cells and supports the possibility that residues at positions 71, 74, and 139 of the 10A1 envelope SU protein account for the expanded host range of 10A1.

Structure-function studies of the human immunodeficiency virus type 1 matrix protein, p17.

The human immunodeficiency virus type 1 (HIV-1) matrix protein, p17, plays important roles in both the early and late stages of the viral life cycle. Using our previously determined solution structure of p17, we have undertaken a rational mutagenesis program aimed at mapping structure-function relationships within the molecule. Amino acids hypothesized to be important for p17 function were mutated and examined for effect in an infectious proviral clone of HIV-1. In parallel, we analyzed by nuclear magnetic resonance spectroscopy the structure of recombinant p17 protein containing such substitutions. These analyses identified three classes of mutants that were defective in viral replication: (i) proteins containing substitutions at internal residues that grossly distorted the structure of recombinant p17 and prevented viral particle formation, (ii) mutations at putative p17 trimer interfaces that allowed correct folding of recombinant protein but produced virus that was defective in particle assembly, and (iii) substitution of basic residues in helix A that caused some relocation of virus assembly to intracellular locations and produced normally budded virions that were completely noninfectious.

Functional analysis of the cytoplasmic tail of Moloney murine leukemia virus envelope protein.

The cytoplasmic tail of the immature Moloney murine leukemia virus (MoMuLV) envelope protein is approximately 32 amino acids long. During viral maturation, the viral protease cleaves this tail to release a 16-amino-acid R peptide, thereby rendering the envelope protein fusion competent. A series of truncations, deletions, and amino acid substitutions were constructed in this cytoplasmic tail to examine its role in fusion and viral transduction. Sequential truncation of the cytoplasmic tail revealed that removal of as few as 11 amino acids resulted in significant fusion when the envelope protein was expressed in NIH 3T3 cells, similar to that seen following expression of an R-less envelope (truncation of 16 amino acids). Further truncation of the cytoplasmic tail beyond the R-peptide cleavage site toward the membrane-spanning region had no additional effect on the level of fusion observed. In contrast, some deletions and nonconservative amino acid substitutions in the membrane-proximal region of the cytoplasmic tail (residues L602 to F605) reduced the amount of fusion observed in XC cell cocultivation assays, suggesting that this region influences the fusogenicity of full-length envelope protein. Expression of the mutant envelope proteins in a retroviral vector system revealed that decreased envelope-mediated cell-cell fusion correlated with a decrease in infectivity of the resulting virions. Additionally, some mutant envelope proteins which were capable of mediating cell-cell fusion were not efficiently incorporated into retroviral particles, resulting in defective virions. The cytoplasmic tail of MoMuLV envelope protein therefore influences both the fusogenicity of the envelope protein and its incorporation into virions.

Murine leukemia virus-based Tat-inducible long terminal repeat replacement vectors: a new system for anti-human immunodeficiency virus gene therapy.

We have constructed new murine leukemia virus (MLV)-based vectors (TIN vectors) which, following integration, contain human immunodeficiency virus (HIV) type 1 U3 and R sequences in place of the MLV U3 and R regions. This provides, for the first time, single transcriptional unit retroviral vectors under the control of Tat. TIN vectors have several advantages for anti-HIV gene therapy applications.

Conserved sequences in the carboxyl terminus of integrase that are essential for human immunodeficiency virus type 1 replication.

We have previously identified a residue in the carboxyl terminus of human immunodeficiency virus type 1 integrase (HIV-1 IN), W-235, the requirement for which is only revealed in viral assays for integrase function (P. M. Cannon, W. Wilson, E. Byles, S. M. Kingsman, and A. J. Kingsman, J. Virol. 68:4768-4775, 1994). Our further analysis of this region of retroviral IN has now identified several sequence motifs which are conserved in all the retroviruses we examined, apart from human spumaretrovirus. We have made mutations within these motifs in HIV-1 IN and examined their phenotypes when reintroduced into an infectious proviral clone. The deleterious effects of several of these mutations demonstrate the importance of these regions for IN function in vivo. We observed a further discrepancy, at a motif that is only conserved in the lentiviruses, in the ability of mutants to function in in vitro and in vivo assays. Substitutions both in this region and at W-235 abolish HIV-1 infectivity but do not affect particle production, morphology, reverse transcription, or nuclear import in T-cell lines. Taken together with the in vitro data suggesting that neither of these residues is directly involved in the catalytic reactions of IN, it seems likely that we have identified regions of IN that are essential for interactions with other components of the integration machinery.

Oestrogen and progesterone receptors as prognostic variables in hormonally treated breast cancer.

This study directly compares ER status and PgR status of primary tumour tissue measured by enzyme immunoassays for prediction of response to therapy and survival in 99 women with breast cancer treated by hormone therapy. ER and PgR status alone both correlated with response to therapy (p = 0.002 and p = 0.02 respectively), time to progression (p < 0.0001 and p = 0.003 respectively) and survival (p < 0.001 and p = 0.01 respectively). 67% of tumours ER(+)/PgR(+) showed responsive or static disease compared to 25% of tumours ER(-)/PgR(-). Tumours of mixed phenotype (i.e. ER(+)/PgR(-) and ER(-)/PgR(+)) showed an intermediate response rate of 46%. Similar findings were observed when tumour phenotype was compared with overall survival. Combining ER and PgR allows more accurate prediction of clinical outcome but does not aid in selecting individual patients for endocrine therapy.

A transient three-plasmid expression system for the production of high titer retroviral vectors.

We have constructed a series of MLV-based retroviral vectors and packaging components expressed from the CMV promoter and carried on plasmids containing SV40 origins of replication. These two features greatly enhanced retroviral gene expression when introduced into cell lines carrying the SV40 large T antigen. The two packaging components, gag-pol and env, were placed on separate plasmids to reduce helper virus formation. Using a highly transfectable human cell line and sodium butyrate to further increase expression of each component, we achieved helper-free viral stocks of approximately 10(7) infectious units/ml by 48 h after transient co-transfection with the three plasmid components. This system can be used both for the generation of high titer retroviral stocks for transduction and for the rapid screening of a large number of MLV gag-pol or env mutants.

Human immunodeficiency virus type 1 integrase: effect on viral replication of mutations at highly conserved residues.

Sequence comparisons of the integrase (IN) proteins from different retroviruses have identified several highly conserved residues. We have introduced mutations at 16 of these sites into the integrase gene of human immunodeficiency virus type 1 and analyzed the phenotypes of the resulting viruses. The viruses were all normal for p24 content and reverse transcriptase activity. In addition, all of the mutants could infect T-cell lines and undergo reverse transcription, as assessed by PCR analysis. Most of the mutant viruses also had normal Western blot (immunoblot) profiles, although three of the mutations resulted in reduced signals for IN relative to the wild type on the immunoblots and mutation of residue W235 completely abolished recognition of the protein by pooled sera from human immunodeficiency virus type 1-positive patients. Mutations that have previously been shown to abolish activity in in vitro studies produced noninfectious viruses. The substitution of W235 was notable in producing a noninfectious virus, despite previous reports of this residue being nonessential for IN activity in vitro (A.D. Leavitt, L. Shiue, and H.E. Varmus, J. Biol. Chem. 268:2113-2119, 1993). In addition, we have identified four highly conserved residues that can be mutated without any affect on viral replication in T-cell lines.

Expression of tumour-associated antigens in breast cancer primary tissue compared with serum levels.

The level of expression of six breast cancer-associated antigens or markers (CEA, NCRC-11, HMFG1, HMFG2, D8 and DF3) in primary tumour tissue and patients' serum has been compared. Two-hundred-and-forty-five consecutive patients with newly-diagnosed breast cancer had tumour biopsy and serum samples taken prior to any anti-cancer therapy (Stage I and II disease n = 100, Stage III disease n = 60, Stage IV disease n = 85). No correlations were found between the level of tumour tissue staining and serum levels of the same antigens. Those samples from patients with Stage IV disease were also analysed separately but again no correlations were observed. Therefore the staining pattern of the primary tumour for these antigens is not a predictor of which tumour marker will be elevated in the serum.