JOURNAL OF VIROLOGY, Dec. 2002, p. 11801–11808 0022-538X/02/$04.00⫹0 DOI: 10.1128/JVI.76.23.11801–11808.2002 Copyright © 2002, American Society for Microbiology. All Rights Reserved.
Vol. 76, No. 23
Amino Acid Residues in the Carboxy-Terminal Region of Cottontail Rabbit Papillomavirus E6 Influence Spontaneous Regression of Cutaneous Papillomas Jiafen Hu,1 Nancy M. Cladel,1 Martin D. Pickel,1 and Neil D. Christensen1,2* Department of Pathology, The Jake Gittlen Cancer Research Institute,1 and Department of Microbiology and Immunology,2 College of Medicine, Pennsylvania State University, Hershey, Pennsylvania 17033 Received 24 May 2002/Accepted 21 August 2002
Previous studies have identified two different strains of cottontail rabbit papillomavirus (CRPV) that differ by approximately 5% in base pair sequence and that perform quite differently when used to challenge New Zealand White (NZW) rabbit skin. One strain caused persistent lesions (progressor strain), and the other induced papillomas that spontaneously regressed (regressor strain) at high frequencies (J. Salmon, M. Nonnenmacher, S. Caze, P. Flamant, O. Croissant, G. Orth, and F. Breitburd, J. Virol. 74:10766-10777, 2000; J. Salmon, N. Ramoz, P. Cassonnet, G. Orth, and F. Breitburd, Virology 235:228-234, 1997). We generated a panel of CRPV genomes that contained chimeric and mutant progressor and regressor strain E6 genes and assessed the outcome upon infection of both outbred and EIII/JC inbred NZW rabbits. The carboxy-terminal 77-amino-acid region of the regressor CRPV strain E6, which contained 15 amino acid residues that are different from those of the equivalent region of the persistent CRPV strain E6, played a dominant role in the conversion of the persistent CRPV strain to one showing high rates of spontaneous regressions. In addition, a single amino acid change (G252E) in the E6 protein of the CRPV progressor strain led to high frequencies of spontaneous regressions in inbred rabbits. These observations imply that small changes in the amino acid sequences of papillomavirus proteins can dramatically impact the outcome of natural host immune responses to these viral infections. The data imply that intrastrain differences between separate isolates of a single papillomavirus type (such as human papillomavirus type 16) may contribute to a collective variability in host immune responses in outbred human populations. The cottontail rabbit papilloma virus (CRPV) is a small DNA virus that induces papillomas on hair-bearing skin of both wild and domestic rabbits (reviewed in references 8 and 10). CRPV-induced papillomas can either progress to invasive carcinomas or systemically regress in a proportion of infected animals. These events parallel the outcome of human papillomavirus type 16 (HPV-16) infections in patient populations (19, 22). Several CRPV early genes including the E5, E6, E7, and E8 genes have been reported to be oncogenes (16, 17, 31). However, of these genes, only the E6 and E7 genes are essential for papilloma formation (5, 31, 48). The E8 gene impacts the rate of papilloma growth in both inbred and outbred rabbits (24), while the E5 gene appears to be dispensable for papilloma formation on rabbits (5, 31; our unpublished data). Because the process of malignant conversion is marked by permanent up-regulation of transcripts for the viral oncogenes encoding E6 and E7, these two gene products are considered to be the best viral antigen candidates for host immunity to papillomavirus-associated cancers (reviewed in references 12, 27, 28, and 34). The CRPV E6 gene open reading frame is unique among those of papillomaviruses because of its large size and ability to encode two E6 proteins, identified as long E6 (LE6) and short E6 (SE6) proteins, respectively (2). The LE6 and SE6 coding
* Corresponding author. Mailing address: Department of Microbiology and Immunology, College of Medicine, Pennsylvania State University, Hershey, PA 17033. Phone: (717) 531-6185. Fax: (717) 5316185. E-mail: [email protected]
sequences are located in the same open reading frame but are initiated from different ATG codons. Unlike the E6s of HPVs, none of the CRPV E6 proteins has been reported to bind to either the p53 protein or E6 binding protein (31, 41). Several factors are involved in papilloma evolution, including the host genetic background and viral genetic variability (6). In humans, certain major histocompatibility complex class II (MHC-II) alleles have been reported to be correlated with HPV progression (1, 4, 6, 26). In domestic rabbits, DQ␣ and DR␣ alleles are linked to regression and malignant conversion of CRPV-induced papillomas (15). High rates of regression of CRPV-induced infections occurred in one inbred strain of rabbit but not in another when challenged with the same strain of CRPV (39). In addition, two different subtypes or strains of CRPV from a CRPV virus stock that displayed partial regression in rabbits with a particular MHC-II haplotype have been identified and sequenced (40). We have independently identified and cloned genetically similar progressor (Hershey CRPVp or H.CRPVp) and regressor (Hershey CRPVr or H.CRPVr) strains of CRPV that were collected from papillomas of wild cottontail rabbits from Kansas and Colorado, respectively. H.CRPVp-induced papillomas showed very low regression rates (⬍5%; see Tables 1 and 2) (unpublished observations), whereas H.CRPVr-induced papillomas showed high regression rates of up to 70% in domestic rabbits (unpublished data). The immunological mechanisms by which these strains induced papillomas that show altered regression rates are unknown. Strong evidence indicates that regression is accomplished by host infiltrating immune cells (18, 36, 42) in-
HU ET AL.
TABLE 1. Papilloma evolution following infection with CRPV genomes containing different chimeric E6 gene mutant constructs on inbred and outbred NZW rabbits Expt
Construct (n )
Regressiona (n) in rabbits: Outbred
Regression rate (%) (Pf) in rabbits: Outbred
H.CRPVp 2/24 (4) 2/18 (3) 8.3 11.1e H.CRPVp-E6r (3) 10/14 12/18 71 (0.007) 67 (0.026) H.CRPVp-CE6r (3) 8/16 12/16 50 (0.035) 75 (0.023) 6/6 ND 100 (0.030) H.CRPVp-CE6r 2 (2) NDc ND 0 (1.0) ND H.CRPVp-NE6r (2) 0/6d
H.CRPVp (4) H.CRPVp-CE6r (3)
13 56 (0.195)
20e 67 (0.117)
1 ⫹ 2 H.CRPVp (8) H.CRPVp-CE6r (6)
10 52 (0.012)
15e 71 (0.003)
Number of regressed sites/number of papilloma sites. n, number of rabbits. ND, not determined. d Two papillomas became malignant at 15 months after DNA inoculation. e P ⫽ 1.0 versus-outbred control rabbits by Fisher’s exact test. f P value for the comparison with the corresponding H.CRPVp control group. b c
cluding both CD4⫹ and CD8⫹ T cells (23, 43). The viral protein antigens that are targeted by the infiltrating T cells are unknown. In the present study, we conducted a genetic analysis of the regions of the CRPV E6 gene that influenced spontaneous regression of CRPV-induced papillomas. A set of chimeric and mutant E6 genes were prepared and placed into the H.CRPVp genome, and the resulting strains were used to challenge New Zealand White (NZW) outbred and EIII/JC inbred rabbits. Our results showed that the carboxy terminus-encoding region of the E6 gene, encompassing about 200 bp, was critical for determining papilloma regression. When this region of H.CRPVr was placed into the H.CRPVp genome, high levels of spontaneous regressions were observed. In addition, we found that a single amino acid change (G252E) in the E6 protein of H.CRPVp induced high spontaneous regression rates in EIII/JC inbred rabbits. These observations imply that small changes in the amino acid sequences of papillomavirus
TABLE 2. Papilloma evolution following challenge with constructs encoding single and double amino acid mutations in E6 of H.CRPVp in outbred rabbitsa No. of: Construct
H.CRPVp G200D G222R Q225L S233R G239D⫹D240G G252E D258G D258G⫹P259S P259S EFRdel a
Challenge Regression Persistent Malignant sites sites sites sites (time to cancer [mo])
12 6 6 6 6 6 24 6 6 6 6
0 0 0 0 0 0 0 0 0 0 0
P221L was constructed but not tested.
12 6 6 6 6 6 24 6 6 6 6
4 (ⱖ13) 0 2 (13) 0 1 (13) 0 2 (ⱖ13) 2 (8) 0 1 (8) 1 (13)
FIG. 1. Design of chimeric CRPV E6 genes used in this study (constructs C to E). All CRPV chimeric E6 genes were placed into the H.CRPVp genome after removal of the whole E6p gene from the introduced SacII site to the EcoRI site. All constructs were sequenced prior to infection of rabbit skin.
proteins can dramatically impact the outcome of natural host immune responses to these viral infections. MATERIALS AND METHODS Preparation of plasmids containing mutant CRPV genomes. CRPV genomes were cloned into pUC19 at the SalI site as previously described (25). H.CRPVp and H.CRPVr E6 gene constructs were cloned into a modified pUC19 vector from which the EcoRI site had been eliminated. Chimeric E6 genes were generated by replacement of the region encoding the N terminus (SacII/AvrII) or C terminus (AvrII/EcoRI) of H.CRPVp E6 or the whole H.CRPVp E6 with the equivalent fragments of H.CRPVr (Fig. 1). Point mutations and deletions were introduced into the E6 gene with the QuikChange site-directed mutagenesis kit (Stratagene, La Jolla, Calif.) by following the manufacturer’s instructions. The E6 mutants prepared and tested in this study are summarized in Fig. 2. The chimeric and/or mutated E6 genes were placed back into the H.CRPVp genome, which contained a SacII site at the start of the E6 gene (bp 169; E6 amino acid 6) that was introduced by site-directed mutagenesis for cloning purposes. To generate additional changes in chimeric E6 genes, H.CRPVp CE6r in pUC19 was used as the template for site-directed mutagenesis. Each mutated chimeric E6 gene was replaced back into the H.CRPVp backbone by using the SacII/ EcoRI sites. All CRPV genomes containing mutant and chimeric E6 genes were sequenced from the SacII site to the EcoRI site to confirm the introduced mutations. The constructs were prepared by using cesium chloride density gradient purification and adjusted to a final concentration of 200 g/ml for rabbit skin inoculations. Inoculation of rabbit skin with plasmid viral DNA. NZW outbred rabbits were purchased from Covance (Denver, Pa.), and EIII/JC inbred rabbits were bred and maintained in the animal facilities of the Pennsylvania State University College of Medicine. Prior to viral DNA challenge, the backs of the rabbits were shaved and painted with a mixture of turpentine and acetone (50:50 [vol/vol]) four times in total, once every other day, to make the skin hyperplastic (13, 25). Viral DNA constructs (10 g per site) were placed onto scarified sites in 50-l volumes. All animal care and handling procedures were approved by the Institutional Animal Care and Use Committee of the Pennsylvania State University. Papilloma size determination and statistical analysis. Papillomas were measured in three dimensions (length, width, and height) in millimeters, from which a geometric mean diameter (GMD) was calculated. Measurements were conducted weekly, beginning from 3 weeks after initial viral DNA challenge. Data were represented as the means ⫾ standard errors of the means of GMDs for
VOL. 76, 2002
EFFECTS OF CRPV E6 MUTATIONS ON PAPILLOMA REGRESSION
FIG. 2. Amino acid alignment between H.CRPVp and H.CRPVr E6 proteins over the carboxy-terminal 77 amino acids (encoded by the sequence between the AvrII site and the end of E6 gene). There are 12 aa differences (shaded) between the two E6 proteins over this region. In addition, the H.CRPVr E6 has three fewer amino acids at the C terminus than the H.CRPVp E6 because of a mutation that introduces an early stop codon in the associated gene.
individual papillomas. Statistical significance was determined by unpaired-t-test comparisons. Regression frequencies for papillomas due to strains with mutant and hybrid CRPV genomes were compared to regression rates of H.CRPVpinduced papillomas by using Fisher’s exact probability test for small samples and by the chi-square test.
RESULTS The carboxy-terminal region of H.CRPVr E6 is critical for papilloma regression in both outbred and inbred rabbits. In our initial studies, we tested various chimeric E6 genes engineered into the H.CRPVp strain (Fig. 1 and Table 1). Three, three, and four NZW outbred rabbits were challenged with chimeric H.CRPVp-E6r, H.CRPVp-CE6r, and H.CRPVp (containing the introduced SacII site) constructs at 14, 16, and 24 sites, respectively. Three sites on each of two rabbits were challenged with H.CRPVp-NE6r. Papillomas developed on all sites at about 3 weeks postchallenge. However, papillomas initiated with H.CRPVp-E6r and H.CRPVp-CE6r began to regress around the fifth week. Ten of 14 sites inoculated with H.CRPVp-E6r regressed within 8 weeks, while 8 of 16 sites inoculated with H.CRPVp-CE6r regressed. Two sites challenged with wild-type H.CRPVp regressed around the 10th week. None of the H.CRPVp-NE6r-inoculated sites regressed although growth rates were slower than those for control H.CRPVp sites, and two of the papillomas progressed to squamous cell skin cancers 15 months after initial viral DNA inoculation (Fig. 3A and Table 1). We also conducted experiments to determine whether papillomas induced with these chimeric constructs in EIII/JC inbred rabbits regressed. Nine rabbits were divided into three groups of three rabbits each; one group was inoculated with H.CRPVp, one was inoculated with H.CRPVp-CE6r, and one was inoculated with H.CRPVp-E6r. Each rabbit was challenged at six scarified sites. Six sites on two inbred rabbits were challenged with H.CRPVp-CE6r#2. For sites receiving H.CRPVp-CE6r, papillomas developed at 16 of 18 challenged sites. At week 4 postchallenge, all four papillomas on one rabbit regressed and five of six papillomas on the second rabbit regressed. At week 6, papillomas on two of three rabbits completely regressed while papillomas on the third rabbit re-
FIG. 3. Papilloma growth curves for E6 mutants in outbred rabbits. (A) Six sites on two rabbits were challenged with control and H.CRPVNE6r constructs. Papillomas induced by H.CRPV-NE6r were smaller than those induced by wild-type constructs (P ⬍ 0.05, t test). SE, standard error. (B) Six sites on two rabbits were challenged with control and H.CRPVp-E6EFRdel constructs. There was no difference in papilloma growth rates for these two constructs (P ⬎ 0.05, t test).
HU ET AL.
mained persistent. For the group challenged with H.CRPVpE6r, all challenge sites grew papillomas at week 3. At week 4, all six papillomas on one rabbit regressed, and, by week 7, all six papillomas on the second rabbit had regressed. Papillomas on the third rabbit remained persistent. For the rabbits challenged with H.CRPVp-CE6r#2, all six papillomas regressed around week 7. In contrast, papillomas generated from wildtype H.CRPVp grew progressively with no regressions of any sites until weeks 9 and 10, at which time, two sites on one rabbit regressed (Table 1). The mean sizes of the remaining persistent papillomas on H.CRPVp-E6r- and H.CRPVp-CE6r-infected rabbits were similar to that of papillomas growing at sites challenged with H.CRPVp (data not shown; P ⬎ 0.05). These results demonstrated that the carboxy-terminal portion of H.CRPVr E6 plays a dominant role in spontaneous regression. A second experiment confirmed high rates of regression of H.CRPVp-CE6r-infected sites on both inbred and outbred rabbits (Table 1). Individual amino acid changes in H.CRPV E6 induce persistent and malignant papillomas in outbred rabbits. We have shown above that the carboxy-terminal portion of H.CRPVr E6 is critical for papilloma regression. When protein alignments between H.CRPVp E6 and H.CRPVr E6 were conducted, we found that there were a total of 15 amino acid differences within this C-terminal region of E6 (Fig. 2). Included in these differences are the loss of three amino acids (EFR) in H.CRPVp CE6r due to a point mutation that introduces a stop codon. To investigate which of these 15 amino acids influences regression, we generated a set of single and double amino acid changes into H.CRPVp E6 to match the amino acid(s) present in the H.CRPVr E6 (Fig. 2). Each mutant was used to challenge six scarified sites on two outbred rabbits per construct. H.CRPVp was included as a positive control for the induction of persistent papillomas and was used to challenge two additional sites on each rabbit. All sites challenged with the viral DNA constructs developed papillomas around 3 weeks postchallenge. Papilloma sizes of all constructs were similar (P ⬎ 0.05, t test; Fig. 3B and 4) with the exception of those due to construct G252E, which induced larger papillomas (see Fig. 7A; P ⬍ 0.05, t test). We conducted a second experiment using the H.CRPVp-E6 G252E and control H.CRPVp genomes on three more outbred rabbits and obtained similar results (data not shown). All mutants with single amino acid changes in E6 grew persistent papillomas, some of which progressed to malignancy 13 months later (Table 2). Therefore, single amino acid changes in E6 from H.CRPVr were unable to reverse the progressive phenotype of H.CRPVp when strains expressing mutant versions of E6 were used to challenge outbred rabbits. These data suggest that multiple amino acids within this C-terminal region of E6 (from a total of 15) are required for induction of increased regression of CRPV in outbred rabbits. Multiple amino acid changes induce papilloma regression in outbred but not inbred rabbits. Previous results showed that single amino acid changes within the C-terminal region of H.CRPVp E6 had little impact on the persistent nature of H.CRPVp-induced papillomas on outbred rabbits (Fig. 4). We were interested, therefore, in whether two or more amino acid changes in this region could be sufficient to trigger host immunity leading to regression in outbred rabbits. Additional mu-
FIG. 4. Papilloma growth curve following infection with CRPV genomes encoding single amino acid changes in H.CRPVp E6. Six sites on two rabbits were challenged with each mutant. No significant difference between the sizes of papillomas produced by these constructs and the sizes of those produced by the control H.CRPVp genome were observed (P ⬎ 0.05, t test). SE, standard error.
tants were constructed by using the chimeric H.CRPVp-CE6r construct as the starting template for mutagenesis. The rationale for choosing this construct was that it induced a high proportion of regressions in both inbred and outbred rabbits (Table 1) and encoded only 15 amino acid (aa) residues of the regressor E6 protein that were different from those of the persistor E6 protein (Fig. 2). Because these 15 aa residues, encoded by the CE6r gene, are found among the regressor E6 residues, the mutations are reversed (e.g., G252E is now a back mutation designated E252G) from those shown in Table 2. The mutants that were prepared are illustrated in Fig. 2. All the mutants were challenged at two sites on each of four outbred rabbits (eight sites in total per construct). All the challenge sites developed papillomas at 3 weeks postchallenge. Some of the papillomas began to regress around week 5. At the end of the monitoring period, the mutant H.CRPVpCE6r (G258D⫹S259P⫹D265N) showed regression at six of eight sites. The mutants D239G⫹G240D⫹D265N, D265N, R233S⫹G258D⫹S259P, and R233S showed regressions at five of eight, four of eight, four of eight, and three of six sites, respectively (Table 3). One rabbit challenged with constructs R233S, R233S⫹E252G, D239G⫹G240D, E252G, and G258D⫹ S259P showed natural spontaneous regression of all papillomas, and data from this rabbit were excluded from final calculations of regression rates for these constructs. Thus only six sites on each of three rabbits were available for determinations of regression rates. For them, E252G, R233S⫹E252G, and G258D⫹S259P showed no regressions and the papillomas growing at R233S⫹E252G and G258D⫹S259P sites were larger than those at nonregressed sites induced from the control construct H.CRPVp-CE6r (Fig. 5; P ⬍ 0.05, t test). To determine whether these mutants showed similar phenotypes when used in challenges of EIII/JC inbred rabbits, two persistent mutants (E252G and R233S⫹E252G) and two regressive mutants (R233S and G258D⫹S259P⫹D265N) were tested on four rabbits for a total of eight challenge sites per
VOL. 76, 2002
EFFECTS OF CRPV E6 MUTATIONS ON PAPILLOMA REGRESSION
TABLE 3. Papilloma evolution following challenge with constructs encoding E6 mutations in the CE6r region of the H.CRPVp-CE6r genome on outbred and inbred rabbits No. of indicated sites in: Construct
H.CRPVp-CE6r R233S E252G D265N R233S⫹D265N R233S⫹E252G D239G⫹G240D G258D⫹S259P E252G⫹D265N R233S⫹G258D⫹S259P D239G⫹G240D⫹D265N G258D⫹S259P⫹D265N a
16 6 6 8 8 6 6 6 8 8 8 8
8 3 0 4 2 0 2 0 2 4 5 6
18 8 8 NDa ND 8 ND ND ND ND ND 8
12 0 0 0
ND, not determined.
construct. All the challenge sites grew papillomas around 4 weeks postchallenge, and these sites grew persistently without regression. Papilloma sizes in both H.CRPVp-CE6r R233S⫹E252G and H.CRPVp-CE6r G258D⫹S259P⫹D265N groups were significantly larger than those of the control group (Fig. 6; P ⬍ 0.05, t test). These data demonstrated that there were clear differences in host responses to the mutant CRPV genomes in inbred versus outbred rabbits. A single amino acid change in E6 induces papilloma regression in EIII/JC inbred rabbits. We next examined the growth rate and behavior of the mutant H.CRPVp-E6G252E on EIII/ JC inbred rabbit back skin. Eight inbred rabbits were separated into two groups, and each rabbit was challenged at six scarified sites with either H.CRPVp-E6G252E or H.CRPVp viral DNA
FIG. 6. Growth rates of papillomas induced by H.CRPVp-CE6r and H.CRPVp-E6 control genomes delivered onto inbred rabbit skin. Eight sites on four rabbits were challenged with two constructs. H.CRPVp-CE6rR233S⫹E252G and H.CRPVp-CE6rG258D⫹S259P⫹ D265N induced papillomas significantly bigger than those induced by the control genomes (P ⬍ 0.05, t test), whereas no significant difference in papilloma size was found at sites challenged with H.CRPVp-CE6r R233S and H.CRPVp-CE6r E252G(P ⬎ 0.05, t test). SE, standard error.
(10 g of DNA/per site). Papillomas appeared on both groups at 3 weeks. At week 5, two sites challenged with the H.CRPVpE6G252E mutant DNA began to regress. Regressions of sites inoculated with H.CRPVp-E6G252E continued over the following 5 weeks, such that, by 12 weeks postchallenge, 22 of 24 (92%) sites had regressed. The two remaining sites on one rabbit were persistent but very small compared to those on rabbits infected with the wild-type H.CRPVp DNA (Fig. 7B). None of the control group sites regressed until week 14, when a single site (1 of 12 or 8.3%) regressed (Table 4). Interestingly, H.CRPVp-E6G252E-challenged sites on outbred rabbits (Table 2) did not regress, and papillomas grew to a larger size than those at sites challenged with control H.CRPVp (Fig. 7A). These results indicated that the amino acid change G252E was crucial for regression in certain genetic backgrounds. DISCUSSION
FIG. 5. Papilloma growth curves following infection with H.CRPVpCE6r genomes encoding single or double amino acid changes in CE6r. Eight sites on four rabbits were challenged with each mutant CRPV. Sites challenged with H.CRPVp-CE6rG258D⫹S259P and H.CRPVpCE6rR233S⫹E252G induced bigger papillomas (P ⬍ 0.05, t test) than sites challenged with control H.CRPVp-CE6r, whereas there was no significant difference in papilloma size for sites infected with H.CRPVp-CE6rE252G. SE, standard error.
Intratype variability between two CRPV strains has been reported, and the E6/E7/upstream regulatory region has been identified as an important region for the regression of CRPV infections on rabbits (39). Our study provides additional and new evidence that the carboxy-terminal portion of E6 is the most important region influencing spontaneous regression of CRPV-induced papillomas in both inbred and outbred rabbits. Most striking was the observation that a single amino acid change in E6 (G252E) led to papilloma regression in EIII/JC inbred rabbits. The independent isolation and cloning of these two strains of CRPV (described in this study as progressor strain H.CRPVp and regressor strain H.CRPVr) provided us with a unique set of reagents to study the interaction between viruses and host immunity (39, 40). Our preliminary experiments showed that a chimeric CRPV genome consisting of H.CRPVp-E6r and H.CRPVp-CE6r induced papilloma re-
HU ET AL.
J. VIROL. TABLE 4. Papilloma regression following infection with H.CRPVpG252E genomes on EIII/JC inbred NZW rabbits Expt
1 2 1⫹2
Regression rate (%)
H.CRPVp (4) H.CRPVpG252 E (4) H.CRPVp (4) H.CRPVpG252 E (4) H.CRPVp (8) H.CRPVpG252 E (8)
2/24 22/24 2/16 8/16 4/40 30/40
8.3 92c 12.5e 50d 10 75f
a Number of regressed sites/number of papilloma sites. No regression in 24 challenge sites on H.CRPVpG252E-challenged outbred rabbits was found (see Table 2). b n, number of rabbits. c P ⬍ 0.001 versus value for H.CRPVp by Fisher’s exact test. d P ⫽ 0.052 versus value for H.CRPVp by Fisher’s exact test. e P ⫽ 1.0 versus value in experiment 1 for H.CRPVp by Fisher’s exact test. f P ⬍ 0.001 versus value for H.CRPVp by the chi-square test.
FIG. 7. Outgrowth and regression of H.CRPVp-E6 G252E-induced papillomas. (A) Six sites on two outbred rabbits were challenged with H.CRPVp-E6G252E and H.CRPVp wild-type constructs. Papillomas induced by the mutant construct were larger than those induced by the control construct in outbred rabbits (P ⬍ 0.05, t test). SE, standard error. (B) Twenty-four sites on four inbred rabbits were challenged with mutant H.CRPVp-E6G252E and wild-type constructs, respectively. Twenty-two of 24 sites regressed at the end of this experiment. The two remaining papillomas were smaller than those at sites challenged with control H.CRPVp (P ⬍ 0.01, t test).
gression in rabbits. DNA sequence alignments indicated that our H.CRPVp genome was highly homologous with that of the published prototype (⬎99%) and that the encoded E6 proteins were identical with the exception of a single amino acid difference at residue 81 (from Pro to Thr). H.CRPVr E6 was identical to the published regressor CRPV strain (40). Collectively, our results demonstrated that H.CRPV E6r, particularly the carboxy-terminal region, was critical for triggering papilloma regression. To determine which amino acids in CRPV E6 lead to papilloma regression, single amino acid mutations and deletions were introduced into H.CRPVp E6. Introduced amino acid changes were selected to match the positionally equivalent residue found in H.CRPVr E6. We have successfully challenged rabbits with plasmids containing the H.CRPVp genome, and almost all of the challenge sites grew papillomas with very low rates of spontaneous regression in outbred rabbits (⬍5% [unpublished observations]). All
strains having genomes encoding versions of E6 with single amino acid mutations therefore were used for challenges in the same way as the wild-type H.CRPVp strain. Previous observations have shown that CRPV E6 is critical for the formation of CRPV-induced papillomas (31). All of these mutant CRPV genomes induced papillomas with the same time kinetics of appearance as that of papillomas at sites inoculated with H.CRPVp. These results indicate that all E6 mutations were functional with respect to the induction of papillomas in rabbit skin. However, none of the papillomas induced by these constructs on outbred rabbits showed any regression. One of the E6 mutants with a single amino acid change (H.CRPVpE6G252E) induced papillomas that were significantly larger than those at sites inoculated with wild-type H.CRPVp (Fig. 7A). Structure-function studies of the mutant E6 proteins generated in this study were not conducted. Previous in vitro studies showed that some LE6 deletion mutants had higher transforming activities than the wild-type LE6 (17, 31). The authors suggested that these mutations in LE6 produced an E6 protein that was more accessible to E6 binding factors, thus leading to enhanced cell proliferation (17). We have tested several hybrid and mutant E6 genes for their capacities to alter the proliferation of NIH 3T3 cells in vitro. The E6pG252E gene and the wild-type H.CRPVp E6 gene showed similar transforming activities (data not shown). We have also found that hybrid E6 genes containing portions of CRPV and rabbit oral papillomavirus showed transforming activity in vitro but failed to generate papillomas in vivo (data not shown). These later observations indicate that some in vitro structure-function studies of mutant E6 genes cannot predict the function of E6 within the context of the virus life cycle in vivo. Variable host responses to CRPV and HPV infections have been attributed to host genetics, particularly to certain polymorphic MHC-II molecules (1, 4, 15, 26, 29, 33, 39). However, more recent data have described numerous isolates of certain HPV types such as HPV-16 that represent variants with small sequence differences, producing amino acid changes, scattered throughout the genome (7, 46). Epidemiological studies have indicated that there is a variable host response to these different variants, with a correlation between particular variants and the persistence of infections and increased malignant progression (20, 49, 50). We found that EIII/JC inbred rabbits, when
VOL. 76, 2002
EFFECTS OF CRPV E6 MUTATIONS ON PAPILLOMA REGRESSION
challenged with H.CRPVp DNA, always showed a rate of spontaneous regression slightly higher, but not to a statistically significant degree, than that shown by outbred rabbits (Table 1) (unpublished observations). These observations suggested that host genetic differences between the inbred and outbred rabbits had little impact on the outcome of H.CRPVp infections. Interestingly, the MHC-II DR and DQ alleles of EIII/JC rabbits were rarely observed among the outbred rabbits that were purchased locally. Thus certain differences in host genetics could help explain our data showing that papillomas induced by H.CRPVp-E6G252E spontaneously regressed with high frequencies (92%) in EIII/JC inbred rabbits but persisted in a small group of outbred rabbits. These results agree with published data indicating that MHC-II differences correlate with different host responses to papillomavirus infections (26, 39). Our studies demonstrated that amino acid changes in CRPV E6 led to spontaneous regressions following infection of rabbit skin with the mutant CRPV genomes. We interpreted these findings to indicate that the amino acid changes in E6 generated an increased immune response to E6 within the developing papilloma. Single amino acid changes in processed peptides that bind to the peptide binding pockets of both MHC-I and MHC-II generate profound differences in the affinity for binding to MHC molecules and dramatically alter T-cell-mediated immune responses to particular proteins (3, 14, 47). Previous studies have shown that spontaneous regression of papillomas is preceded by infiltration of both CD4⫹ and CD8⫹ T cells (35, 36, 43). Collectively, these data support the hypothesis that small changes in amino acid sequences in E6 alter host immune responses to CRPV infection. An alternative and possibly contributing outcome of the amino acid changes in E6 to papilloma growth is the alteration of expression levels or of the affinity of the mutant E6 for cellular factors that ultimately impact the E6 (and other) viral protein expression rates. Differences in the behavior of E6 with respect to p53 binding and degradation between natural variants of HPV-16 E6 were observed (46). Within the developing papilloma, therefore, there may be increased proliferation and/or apoptosis of virus-infected cells. Thus, the impact on immune responses could occur as a consequence of increased expression of E6 and/or other viral proteins within virus-infected cells, with subsequent increased immune stimulation of antigen-specific effector cells. Increased apoptosis of papilloma cells may provide another mechanism for increased antigen exposure to immune effector cells since apoptotic tumor cells can be processed for antigen presentation by dendritic cells (21, 38). All these potential mechanisms, however, ultimately favor an increased immune response to papillomas that develop from CRPV with certain mutations in E6. Some more recent studies in our laboratory (J. Hu et al., unpublished data) have shown that treatment of rabbits with the immunosuppressive drug cyclosporine A allows persistent growth of several constructs, such as H.CRPVr and H.CRPVp-E6r, that usually show high rates of regression. We also note that certain mutant CRPV genomes such as that of H.CRPVp-E6G252E induced persistent papillomas on outbred rabbits. Further experiments are under way to assess whether inbred rabbits that show regression of H.CRPVp-E6G252E infections show enhanced immune responses (e.g., skin delayed-type hypersensitivity 
to peptides containing the mutation versus peptides that are equivalent to wild-type H.CRPVp E6 at position 252). We noticed that the rate of regressions of sites challenged with the regressor strains was seldom 100%, even in inbred rabbits. Clearly, there are additional parameters other than host genetics and viral genotypes that contribute to spontaneous regression of CRPV infections. We have observed that growth rates of papillomas are similar but not identical from rabbit to rabbit and from site to site. Further observations include the finding that, once papillomas reach a certain large size (and/or grow very rapidly), there is a much lower chance for spontaneous regressions to occur. This observation suggests that the microenvironment of papillomas may negatively impact host immune responses, as has been observed in developing tumors in animal models and patient populations (9, 37, 44). Many studies have shown that the tumor microenvironment contains factors that down-regulate immune responses and immune effector populations (11, 30, 32, 45). In conclusion, we observed that amino acid changes in the carboxy-terminal portion of E6 of the CRPV progressor strain led to increased spontaneous regressions. Increased spontaneous regressions could occur following a single amino acid change in CRPV E6 in an inbred rabbit line. These data have implications for the outcome of HPV infections in patient populations where natural variants that have small variations in amino acid sequences in E6 are found. ACKNOWLEDGMENTS This work was supported by National Cancer Institute grant RO1 CA47622 from the National Institutes of Health and by the Jake Gittlen Memorial Golf Tournament. We thank John Sundberg (The Jackson Laboratory, Bar Harbor, Maine) for the gift of papilloma tissue taken from an infected cottontail rabbit from Colorado. REFERENCES 1. Apple, R. J., H. A. Erlich, W. Klitz, M. M. Manos, T. M. Becker, and C. M. Wheeler. 1994. HLA DR-DQ associations with cervical carcinoma show papillomavirus-type specificity. Nat. Genet. 6:157–162. 2. Barbosa, M. S., and F. O. Wettstein. 1988. The two proteins encoded by the cottontail rabbit papillomavirus E6 open reading frame differ with respect to localization and phosphorylation. J. Virol. 62:1088–1092. 3. Beekman, N. J., P. A. van Veelen, T. van Hall, A. Neisig, A. Sijts, M. Camps, P. M. Kloetzel, J. J. Neefjes, C. J. Melief, and F. Ossendorp. 2000. Abrogation of CTL epitope processing by single amino acid substitution flanking the C-terminal proteasome cleavage site. J. Immunol. 164:1898–1905. 4. Beskow, A. H., A. M. Josefsson, and U. B. Gyllensten. 2001. HLA class II alleles associated with infection by HPV16 in cervical cancer in situ. Int. J. Cancer 93:817–822. 5. Brandsma, J. L., Z.-H. Yang, D. DiMaio, S. W. Barthold, E. Johnson, and W. Xiao. 1992. The putative E5 open reading frame of cottontail rabbit papillomavirus is dispensable for papilloma formation in domestic rabbits. J. Virol. 66:6204–6207. 6. Breitburd, F., N. Ramoz, J. Salmon, and G. Orth. 1997. HLA control in the progression of human papillomavirus infections. Semin. Cancer Biol. 7:359– 371. 7. Chan, S.-Y., L. Ho, C.-K. Ong, V. Chow, B. Drescher, M. Du ¨rst, J. Ter Meulen, L. Villa, J. Luande, H. N. Mgaya, and H.-U. Bernard. 1992. Molecular variants of human papillomavirus type 16 from four continents suggest ancient pandemic spread of the virus and its coevolution with humankind. J. Virol. 66:2057–2066. 8. Christensen, N. D., R. Han, and J. W. Kreider. 2000. Cottontail rabbit papillomavirus (CRPV), p. 485–502. In R. Ahmed and I. Chen (ed.), Persistent viral infections. John Wiley & Sons Ltd., Sussex, England. 9. Costello, R. T., J. A. Gastaut, and D. Olive. 1999. Tumor escape from immune surveillance. Arch. Immunol. Ther. Exp. 47:83–88. 10. Evans, C. A., A. L. Rashad, and N. K. Mottet. 1964. The papilloma of rabbits induced by the virus of Shope: histologic features related to amount of virus in the tumor, p. 587–600. In W. Montagna and W. C. Lobitz, Jr. (ed.), The epidermis. Academic Press, New York, N.Y.
HU ET AL.
11. Forni, G., P. L. Lollini, P. Musiani, and M. P. Colombo. 2000. Immunoprevention of cancer: is the time ripe? Cancer Res. 60:2571–2575. 12. Frazer, I. H. 1996. Immunology of papillomavirus infection. Curr. Opin. Immunol. 8:484–491. 13. Friedewald, W. F. 1942. Cell state as affecting susceptibility to a virus; Enhanced effectiveness of the rabbit papilloma virus on hyperplastic epidermis. J. Exp. Med. 75:197–220. 14. Garcia-Peydro, M., A. Paradela, J. R. Lamas, and J. A. Lopez De Castro. 1999. Peptide presentation to an alloreactive CTL clone is modulated through multiple mechanisms involving polymorphic and conserved residues in HLA-B27. J. Immunol. 163:6060–6064. 15. Han, R., F. Breitburd, P. N. Marche, and G. Orth. 1992. Linkage of regression and malignant conversion of rabbit viral papillomas to MHC class II genes. Nature 356:66–68. 16. Han, R., N. M. Cladel, C. A. Reed, and N. D. Christensen. 1998. Characterization of transformation function of cottontail rabbit papillomavirus E5 and E8 genes. Virology 251:253–263. 17. Harry, J. B., and F. O. Wettstein. 1996. Transforming properties of the cottontail rabbit papillomavirus oncoproteins LE6 and SE6 and of the E8 protein. J. Virol. 70:3355–3362. 18. Haskill, J. S., M. Key, L. A. Radov, E. Parthenais, J. H. Korn, J. W. Fett, Y. Yamamura, F. DeLustro, J. Vesley, and G. Gant. 1979. The importance of antibody and macrophages in spontaneous and drug-induced regression of the T1699 mammary adenocarcinoma. J. Reticuloendothel. Soc. 26:417–425. 19. Hildesheim, A. 1997. Human papillomavirus variants: implications for natural history studies and vaccine development efforts. J. Natl. Cancer Inst. 89:752–753. 20. Hildesheim, A., M. Schiffman, C. Bromley, S. Wacholder, R. Herrero, A. Rodriguez, M. C. Bratti, M. E. Sherman, U. Scarpidis, Q. Q. Lin, M. Terai, R. L. Bromley, K. Buetow, R. J. Apple, and R. D. Burk. 2001. Human papillomavirus type 16 variants and risk of cervical cancer. J. Natl. Cancer Inst. 93:315–318. 21. Hoffmann, T. K., N. Meidenbauer, G. Dworacki, H. Kanaya, and T. L. Whiteside. 2000. Generation of tumor-specific T-lymphocytes by cross-priming with human dendritic cells ingesting apoptotic tumor cells. Cancer Res. 60:3542–3549. 22. Hopfl, R., K. Heim, N. Christensen, K. Zumbach, U. Wieland, B. Volgger, A. Widschwendter, S. Haimbuchner, E. Muller-Holzner, M. Pawlita, H. Pfister, and P. Fritsch. 2000. Spontaneous regression of CIN and delayed-type hypersensitivity to HPV-16 oncoprotein E7. Lancet 356:1985–1986. 23. Ho ¨pfl, R. M., N. D. Christensen, M. G. Angell, and J. W. Kreider. 1993. Skin test to assess immunity against cottontail rabbit papillomavirus antigens in rabbits with progressing papillomas or after papilloma regression. J. Investig. Dermatol. 101:227–231. 24. Hu, J., R. Han, N. M. Cladel, M. D. Pickel, and N. D. Christensen. 2002. Intracutaneous DNA vaccination with E8 gene of cottontail rabbit papillomavirus induces protective immunity against virus challenge in rabbits. J. Virol. 76:6453–6459 25. Kreider, J. W., N. M. Cladel, S. D. Patrick, P. A. Welsh, S. L. DiAngelo, J. M. Bower, and N. D. Christensen. 1995. High efficiency induction of papillomas in vivo using recombinant cottontail rabbit papillomavirus DNA. J. Virol. Methods 55:233–244. 26. Lin, P., L. A. Koutsky, C. W. Critchlow, R. J. Apple, S. E. Hawes, J. P. Hughes, P. Toure, A. Dembele, and N. B. Kiviat. 2001. HLA class II DR-DQ and increased risk of cervical cancer among Senegalese women. Cancer Epidemiol. Biomark. Prev. 10:1037–1045. 27. Ling, M., M. Kanayama, R. Roden, and T. C. Wu. 2000. Preventive and therapeutic vaccines for human papillomavirus-associated cervical cancers. J. Biomed. Sci. 7:341–356. 28. Lowy, D. R., and J. T. Schiller. 1998. Papillomaviruses and cervical cancer: pathogenesis and vaccine development. J. Natl. Cancer Inst. Monogr. 23:27– 30. 29. Maciag, P. C., N. F. Schlecht, P. S. Souza, E. L. Franco, L. L. Villa, and M. L. Petzl-Erler. 2000. Major histocompatibility complex class II polymorphisms and risk of cervical cancer and human papillomavirus infection in Brazilian women. Cancer Epidemiol. Biomark. Prev. 9:1183–1191. 30. Marincola, F. M., E. M. Jaffee, D. J. Hicklin, and S. Ferrone. 2000. Escape
31. 32. 33. 34. 35. 36. 37. 38. 39.
47. 48. 49.
of human solid tumors from T-cell recognition: molecular mechanisms and functional significance. Adv. Immunol. 74:181–273. Meyers, C., J. Harry, Y.-L. Lin, and F. O. Wettstein. 1992. Identification of three transforming proteins encoded by cottontail rabbit papillomavirus. J. Virol. 66:1655–1664. Mocellin, S., E. Wang, and F. M. Marincola. 2001. Cytokines and immune response in the tumor microenvironment. J. Immunother. 24:392–407. Montoya, L., I. Saiz, G. Rey, F. Vela, and N. Clerici-Larradet. 1998. Cervical carcinoma: human papillomavirus infection and HLA-associated risk factors in the Spanish population. Eur. J. Immunogenet. 25:329–337. Nicholls, P. K., and M. A. Stanley. 2000. The immunology of animal papillomaviruses. Vet. Immunol. Immunopathol. 73:101–127. Okabayashi, M., M. G. Angell, L. R. Budgeon, and J. W. Kreider. 1993. Shope papilloma cell and leukocyte proliferation in regressing and progressing lesions. Am. J. Pathol. 142:489–496. Okabayashi, M., M. G. Angell, N. D. Christensen, and J. W. Kreider. 1991. Morphometric analysis and identification of infiltrating leucocytes in regressing and progressing Shope rabbit papillomas. Int. J. Cancer 49:919–923. Pardoll, D. M. 1998. Cancer vaccines. Nat. Med. 4:525–531. Ronchetti, A., G. Iezzi, M. C. Crosti, M. P. Garancini, M. P. Protti, and M. Bellone. 1999. Role of antigen-presenting cells in cross-priming of cytotoxic T lymphocytes by apoptotic cells. J. Leukoc. Biol. 66:247–251. Salmon, J., M. Nonnenmacher, S. Caze, P. Flamant, O. Croissant, G. Orth, and F. Breitburd. 2000. Variation in the nucleotide sequence of cottontail rabbit papillomavirus a and b subtypes affects wart regression and malignant transformation and level of viral replication in domestic rabbits. J. Virol. 74:10766–10777. Salmon, J., N. Ramoz, P. Cassonnet, G. Orth, and F. Breitburd. 1997. A cottontail rabbit papillomavirus strain (CRPVb) with strikingly divergent E6 and E7 oncoproteins: an insight in the evolution of papillomaviruses. Virology 235:228–234. Schmitt, A., J. B. Harry, B. Rapp, F. O. Wettstein, and T. Iftner. 1994. Comparison of the properties of the E6 and E7 genes of low- and high-risk cutaneous papillomaviruses reveals strongly transforming and high Rb-binding activity for the E7 protein of the low-risk human papillomavirus type 1. J. Virol. 68:7051–7059. Selvakumar, R., L. A. Borenstein, Y. L. Lin, R. Ahmed, and F. O. Wettstein. 1994. T-cell response to cottontail rabbit papillomavirus structural proteins in infected rabbits. J. Virol. 68:4043–4048. Selvakumar, R., A. Schmitt, T. Iftner, R. Ahmed, and F. O. Wettstein. 1997. Regression of papillomas induced by cottontail rabbit papillomavirus is associated with infiltration of CD8⫹ cells and persistence of viral DNA after regression. J. Virol. 71:5540–5548. Seo, N., S. Hayakawa, M. Takigawa, and Y. Tokura. 2001. Interleukin-10 expressed at early tumour sites induces subsequent generation of CD4⫹ T-regulatory cells and systemic collapse of antitumour immunity. Immunology 103:449–457. Sinkovics, J. G., and J. C. Horvath. 2000. Vaccination against human cancers (review). Int. J. Oncol. 16:81–96. Stoppler, M. C., K. Ching, H. Stoppler, K. Clancy, R. Schlegel, and J. Icenogle. 1996. Natural variants of the human papillomavirus type 16 E6 protein differ in their abilities to alter keratinocyte differentiation and to induce p53 degradation. J. Virol. 70:6987–6993. Suhrbier, A., S. J. Rodda, P. C. Ho, P. Csurhes, H. Dunckley, A. Saul, H. M. Geysen, and C. M. Rzepczyk. 1991. Role of single amino acids in the recognition of a T cell epitope. J. Immunol. 147:2507–2513. Wu, X., W. Xiao, and J. Brandsma. 1994. Papilloma formation by cottontail rabbit papillomavirus requires E1 and E2 regulatory genes in addition to E6 and E7 transforming genes. J. Virol. 68:6097–6102. Xi, L. F., L. A. Koutsky, D. A. Galloway, J. Kuypers, J. P. Hughes, C. M. Wheeler, K. K. Holmes, and N. B. Kiviat. 1997. Genomic variation of human papillomavirus type 16 and risk for high grade cervical intraepithelial neoplasia. J. Natl. Cancer Inst. 89:796–802. Zehbe, I., E. Wilander, H. Delius, and M. Tommasino. 1998. Human papillomavirus 16 E6 variants are more prevalent in invasive cervical carcinoma than the prototype. Cancer Res. 58:829–833.