Proc. Nat. Acad. Sci. USAVol. 70, No. 11, pp. 3225-3229, November 1973

Herpes Simplex and Herpes Genitalis Viruses in Etiology of SomeHuman Cancers*

(DNA tumor viruses/nonvirion antigens/complement fixation/HeLa cells/HEp2 cells)

ALBERT B. SABINt AND GIIULIO TARRO$

National Cancer Institute Frederick Cancer Research Center, Fort Detrick, Frederick, Maryland 21701

Contributed by Albert B. Sabin, July 5, 1973

ABSTRACT The results of complement fixation testson 202 sera from people without cancer and from patientswith cancer in 29 different areas of the body indicated thatonly those with nine varieties of advanced cancer (lip,mouth, oropharynx, nasopharynx, kidney, urinary blad-der, prostate, cervix uteri, and vulva-all of 56 tested) gavepositive specific reactions with nonvirion antigens inducedby the DNA herpes simplex (HSV 1) and herpes genitalis(HSV 2) viruses. None of 57 people without cancer (includ-ing 10 with current and 18 with recurrent HSV 1 or HSV 2infections), none of 81 patients with 20 other varieties ofadvanced cancer (gum, tongue, tonsil, salivary gland, ac-cessory sinus, epiglottis, lung-bronchus, stomach, colon,breast, corpus uteri, ovary, testis, liver, thyroid, Wilms'embryonal kidney, melanoma, Hodgkin's disease, acutelymphocytic leukemia, and acute myelocytic leukemia),and none of four women with early malignant changes inthe cervix uteri gave positive results. The seven patientswith advanced cancer of the lip or oropharynx gave posi-tive reactions with HSV 1 but not with HSV 2 nonvirionantigens (compatible with involvement of only HSV 1),all of the 13 women with advanced cancer of the cervixuteri and the one woman with advanced cancer of the vulvagave positive reactions with both HSV 1 and HSV 2 non-virion antigens (compatible with involvement of only HSV2), while among the 35 other positive patients only two(one with cancer of the kidney and one with cancer of thebladder) reacted with HSV 1 and not at all with HSV 2nonvirion antigens. Positive sera failed to react with cellsharvested at different times after high-multiplicity in-fection with the DNA vaccinia virus. Massive absorption ofpositive sera with trypsinized, uninfected human embry-onic kidney cells failed to remove, or lower the titer of,the HSV 1 and HSV 2 nonvirion antibodies.

All of these data taken together are interpreted as in-dicating that HSV 1 and HSV 2 play an etiologic role incertain human cancers, because they provide the kind of

Abbreviations: HSV 1, herpes simplex (labialis) virus, subtype1; HSV 2, herpes simplex (genitalis) virus, subtype 2; GPK,guinea pig kidney; RK, rabbit kidney; HEp2, a human cell linederived from an epidermoid carcinoma of the larynx; PFU,plaque-forming unit.* Preliminary report of data presented at the I10th AnnualMeeting of the National Academy of Sciences, Washington,D.C., April 24, 1973.t This study was conducted while I)r. Sabin was a FogartyScholar-in-Residence, Fogarty International Center for AdvancedStudy in the Health Sciences. Reprint requests should be sentto Dr. Sabin, Fogarty International Center, National Institutesof Health, Bethesda, \Id. 20014.t During this work Dr. Tarro was on leave from, and is currentlyat, Virologia Oncologica, Istituto di Clinica -Medica I, Universitadi Napoli, Polichinico, 80138 Napoli, Italia.

3225

evidence by which virus-free experimental cancers can beproved to have been originally induced by such DNA vi-ruses as polyoma, Simian Virus 40, or certain types ofadenovirus.

The evidence required for establishing DNA viruses, whichcause ordinary infections in animals anid humans, as etiologicor inu(lUcilg agents in cancer is different from that which hasestablished several unique RNA viruses, unassociated withordinary infections as the (cause of various naturally occurringmalignancies of lower animals. Some DNA viruses associatedwith ordinary infections (i.e., l)olyooma of mice, Simian Virus40 of rhesus monkeys, anad some types of human and animaladenoviruses) can produce experimental cancers in which in-fectious virus, 01 infectious DNA, or any of the structural(virion) antigens cannot be detected either dlirectly or bytheir capacity to induce antibody in the tumor-bearing ani-mals. There are in such cancers,howei-er, antigens that arespecific for the virus that induced the exlperimental cancer,and tumor-bearing animals develop specific antibodies forthem (1, 2). These specific tumor antigens are identical tononvirion antigens (i.e., not structural comlpoIients of the as-sembled virus particle) that are synthesized shortly afterhioh-multiplicitv infection of suscel)tible cultured cells. evenwhen synthesis of new DNA is prevented by al)l)roi)riate in-hibitors (2-7). Phenomena associated with experimental DNAvirus cancers were studied in detail with special emphasis onthose manifestations that could be important ill the searchfor evidence of the possible involvement of the well-knownDNA viruses of humans, i.e., maintained in nature only bytransmission from person to person, in human cancers (8-10).The subsequent search for nonvirion antigens in cells infectedwith subtypes 1 and 2 of herpes simplex virus (HSV) andthe development of methods for identifying them, studyingtheir distinctive properties, and differentiating them frompreexisting antigens, whose concentration is increased byboth malignant transformation and various viral infections,have been reported (11-13).The preceding studies provided the necessary reagents and

methodology for testing human sera for HSV nonvirion anti-bodies, as well as the important fact that sera from smallnumbers of randomly selected adults and from people withrecent HSV 1 or HSV 2 infections had no HSV nonvirionantibodies demonstrable by comllement fixation (11, 13).Thus, we could proceed with the final Chase of the plan (8-10)to explore the possible role of HSV 1 and HSV 2 in humancancer by determining whether or not the sera of certain

Dow

nloa

ded

by g

uest

on

Feb

ruar

y 18

, 202

0

3226 Medical Sciences: Sabin and Tarro

TABLE 1. Complement fixation tests wnonvirion antigens and sera of peo

Group te

Current herpesvirus infectionsHSV 1, 4; HSV 2,6.

Recurrent herpetic lesionsLabial, 7; genital (HSV 2), 1 1.

Pregnant womenLast trimester, 9; postpartum, 7.

Random adultsTotal

cancer patients might exhibit a specifinonvirion antigens that would not be ithe special cancer or cancers even amor

age with frequent recurring infections f

MATERIALS AND MI

Cell Lines and Their Cultivation. H]preparation of HSV 1 and HSV 2 virintion of sera and also for preparationVery large quantities of HEp2 cells weLaboratories, Rockville, Md. either fiprescription bottles in 90% minimum Ifetal-bovine serum without antibiotictrypsinized cells which we grew oursmedium (GIBCO powder rehydrateregular calf serum + 0.11% NaHCOa0.22% NaHCO3 later + penicillin ('mycin (0.2 mg/ml), and gentamycinGPK cultures were used only for provirion antigens. Kidneys of 2- to 3-drtrypsinized in warmed 0.25% trypsinsaline (20 ml/g of kidney) and thorolarge quantities of cold growth mediumfuge. The volume of cell pack was dettion at 600 rpm for 10 min, and 40 ml cvolume prepared in warm growth medi

ounce prescription bottle. The growl85% of 0.5% lactalbumin hydrolysatEwithout NaHCO3 (dehydrated prodiIsland, N.Y. 14072) + 5% tryptose pfetal-bovine serum (selected for abse

at 560 for 30 min + 0.035% NaHCO3ml), streptomycin (0.2 mg/ml), and geMedium is changed at 2 days (with3-4 days (0.11% NaHCO3), and atHCO3). Excellent, fully confluent mluse 6-7 days after cells are planted i

scribed is strictly followed. Secondarymonolayers of young (150-200 g) REfor plaque titration of the potency of

multiplicity infection of cells for pantigens.

Viruses Used, Preparation, and St(HSV 1) and Sabin "Justin" (HSV 2)VL strain of vaccinia virus used in t

Proc. Nat. Acad. Sci. USA 70 (1973)

ith HSV 1 and HSV 2 scribed (11-13). Virus stocks were prepared in HEp2 cellsple without cancer in the logarithmic phase of multiplication, which is optimum

for obtaining high yields of virus.No.

positive with Preparation of HEp2 Cells for Absorption of Virion Anti-bodies. We wanted to obtain large numbers of infected cells

No. HSV HSV for storage at 40 to inactivate the nonvirion antigens which inZsted 1 2 HEp2 cells are at peak titers at the end of the infectious pro-

10 0 0 cess (11, 13). Accordingly, densely packed cell sheets, at 7-8days after planting 4 X 106 cells per 32-ounce bottle, are in-

18 0 0 fected with about 5 plaque-forming units (PFU)/cell. Thesame procedure of harvesting, freezing, and sonicating is

16 0 0 followed as for preparation of virus stocks except that the 10%13 0 0 cell suspension is prepared in Eagle's basal medium without57 0 0 serum or phenol red instead of in demineralized water, which

is necessary to avoid marked anticomplementary activity,and storage is at 4° (±0.50) instead of at -80°.

ic reactivity with HSV Preparation of Nonvirion Antigens in GPK and HEp2 Cells.found in the absence of After the medium was decanted from GPK confluent mono-ig people of comparable layers grown for 6-7 days in 32-ounce bottles, 10 ml of Eagle'sor decades. basal medium without serum or phenol red, containing pref-

ETHODS erably about 2 X 108 PFU of HSV, are added. After incuba-ETHDcells used for

tion for 3 hr at 370, the cells are scraped into the medium andonp2 cells were used for centrifuged as for preparation of stock virus, except thaton antigens for absorp- basal medium without serum or phenol red is used to makeof nonvirion antlgens. the 10% suspension. Moreover, after it was frozen andere obtained from Flow thawed, the suspension was not sonicated directly in the cupully grown in 32-ounce but put into a stoppered lusteroid tube which rested in waterEagle's medium + 10% in the cup (because the volume was often too small and alsos, or usually as freshly because that is the way we did it before), and the sonicatedselves in Eagle's basal suspension was frozen uncentrifuged at -80°. For maximumd) + 10% unheated, concentration of nonvirion antigens for complement fixation3 for initial growth and tests, the uncentrifuged suspension was used the following day200 units/ml), strepto- or at most within 4 days. When HEp2 cells were used, they(0.05 mg/ml). Primary were grown and infected as for virus stock and harvested aftereparation of HSV non- 24 hr when the cytopathic effect is complete and the concen-Ly-old guinea pigs were tration of nonvirion antigens is highest (11, 13). The remain-i in phosphate-buffered der of the process was as for GPK cells. An aliquot of the sameughly washed twice with lot of uninfected GPK or HEp2 cells was processed in pre-in a refrigerated centri- ciselv the same manner to provide the control antigen for thetermined by centrifuga- complement fixation test.5f a 1:400 suspension byurn was seeded per 32- Monitoring of Nonvirion and Virion Antisera. Sabinth medium consisted of "Schooler" HSV 1 guinea pig 71 serum (11) and Sabine in Earle's salt solution "Justin" HSV 2 guinea pig 91 serum (13) were used, afteruct of GIBCO, Grand proper virion absorption, to measure the concentration ofhosphate broth + 10% HSV 1 and HSV 2 nonvirion antigens, respectively. Whilence of toxicity) heated HSV 1 nonvirion guinea pig serum detected nonvirion anti-+ penicillin (200 units/ gens induced only by HSV 1 strains, HSV 2 serum detectedntamycin (0.05 mg/ml). nonvirion antigens induced not only by HSV 2 but also by

i 0.035% NaHCO3), at HSV 1 (13). A hyperimmunized rabbit serum prepared by5-6 days (0.22% Na- one of us (A.B.S.) in 1958 against the "Schooler" HSV 1

onolayers are ready for strain, containing an especially broad spectrum of HSVwhen the procedure de- virion antibodies, measured equally well the concentrationr cultures prepared from of virion complement fixation antigen produced by HSV 1 and< cell cultures were used HSV 2.the HSV used for high- Human Sera. Most of the sera from patients with advanced

cancer were provided by Dr. Jack Gruber, Chief, Office ofProgram Resources and Logistics, National Cancer Institute

'orage. Sabin "Schooler" largely from a collection that was made at the M. D. An-strains of HSV and the derson Hospital, Houston, Texas in 1967 and stored at - 700

this work have been de- in aliquot amounts since then. Each of these sera came from

Dow

nloa

ded

by g

uest

on

Feb

ruar

y 18

, 202

0

Herpesviruses 1 and 2 and Human Cancers 3227

a person with histologically confirmed malignancy and evi-dence of metastasis (25 g or more of primary and metastatictumor). Some of the random sera from people without cancerwere from family or other matched controls of these patients.We thank Dr. Brian E. Henderson of the Department ofPathology, University of Southern California School of Medi-cine, Los Angeles, Calif., for the following 13 cancer sera usedin the present tests: two nasopharynx from Chinese patients,one lip, two corpus uteri, five ovary, and three testis. Dr.John L. Sever and Dr. David A. Fucillo, Perinatal ResearchBranch, National Institute of Neurological Diseases andStroke, supplied the sera from women before and after earlymanifestations of malignancy of the cervix uteri, and also thesera from pregnant women without cancer. We especiallythank Dr. A. J. Nahmias, Department of Pediatrics, EmoryUniversity School of Medicine, Atlanta, Ga., for sera on pa-tients with recurrent HSV 2 infections he had been studyingfor many years. These male and female patients ranged inage from 20-64 years (six of them 52 years or older) with re-curring lesions on the penis (two), vulva (two), buttocks(five), thigh (one), and finger (one). Four of these patientshave had recurrent lesions for 15, 15, 20, and 23 years, re-spectively, three with four to six recurrences per year andone with 20-30 per year. Among the others there were fivewith 10-25 recurrences per year. Sera from people with herpeslabialis were received from different sources and were frompeople over 35 years of age.

Absorption of Virion Antibody from Human and Guinea PigSera before Tests for Nonvirion Antibody. The basic proceduresfor absorption were described (3, 11). Although storage atabout 40 had been found to eliminate HSV 1 and HSV 2nonvirion reactivity from infected HEp2 cell suspensions be-tween 7 and 9 days and between 12 and 15 days, respectively(11, 13), when 0.1-ml quantities were being tested, the verymuch larger quantities (100-times or more) required for ab-sorption of the human and guinea pig sera might still containenough nonvirion antigen at 14 days to reduce the titer ofnonvirion antibody in the absorbed sera. Moreover, the orig-inal observations (11, 13) were not made under conditions ofstrict temperature control (an ordinary refrigerator that wasopened many times each day had been used); we could not as-sume that under the conditions of storage used in the presentwork (an infrequently used walk-in refrigerator carefully con-trolled and monitored at 40 + 0.50) the nonvirion antigenswould be inactivated at the same rate. We indeed found thatthe standard guinea pig sera gave much lower nonvirion anti-body titers when they were absorbed with HEp2 cell suspen-sions stored at the strictly controlled 40 temperature for 2-4weeks, while storage for 5-8 weeks yielded higher titers ofnonvirion antibody without loss of capacity to remove thevirion antibody. A mixture of equal parts of HSV 1 and HSV2 antigens stored at +40 was used for absorption of virionantibodies from human sera and concurrently for each testfrom the standard HSV 1 and HSV 2 guinea pig monitoringsera.

Complement Fixation Test. The amount of complement re-quired for the 1.5 exact units to be used in the test was de-termined by a preliminary titration of complement in thepresence of 0.1 ml of the 10% uncentrifuged cell-suspensionantigens and of the absorbed human sera that are diluted 1 :4

plement fixation test (11) was used. Each complement fixationtest always had positive and negative controls: human serafrom people without cancer or with different types of cancer,uninfected cell suspensions representing aliquots of thoseused for preparation of nonvirion antigens and similarly andconcurrently processed and stored, HSV 1 and HSV 2 con-centrated virion antigens, titrations to determine the concen-tration of the nonvirion and virion antigens in the prepara-tions used for the test, titrations of the absorbed monitoringHSV 1 and HSV 2 guinea pig antisera, and, of course, simul-taneous complement controls on all the reagents in the test.The sera in any one test were scrambled and numbered by acode contained in a sealed envelope which was not openeduntil after the results had been recorded.

RESULTS

People Without Cancer. The data in Table 1 show that noneof the sera from 57 people without cancer had nonvirion anti-bodies at the 1:4 dilution tested. The results on the 10 peoplewith current or very recent HSV 1 or HSV 2 infections havebeen reported (11, 13). Available evidence suggests that iiirecurrent HSV lesions, which appear at the same site yearafter year, there is derepression and full expression of totalviral genome which persists in a repressed state in the cells ofthe affected area (14). The completely negative results ob-tained with sera from 18 people with long histories of recur-rent HSV 1 and HSV 2 infections indicate, therefore, thatspecial conditions are required for production and persistenceof antibodies for the nonvirion HSV antigens. The negativeresults obtained with the sera of 16 pregnant women indicatethat the nonvirion HSV antigens we used to test the sera fromcancer patients are free from carcinoembryonic antigens forwhich such sera may have antibodies. The sera from 13random adults included a serum (kindly supplied by Dr. PaulGerber, Bureau of Biologics, Food and Drug Administration,National Institutes of Health, Bethesda, AMd.) that has beenextensively used in studies on human Epsteiin-Barr DNAvirus because of its broad spectrum of antibodies against"early" and "late" antigens of Epstein-Barr virus.

People with Cancer. The data in Table 2 on sera from 137patients with advanced cancer in 29 different sites of the bodyare remarkable because uniformly positive results were ob-tained with 56 sera from patients with the nine listed varietiesof cancer, and uniformly negative results were obtained insimultaneous tests with the same reagents with the sera from81 patients with 20 other varieties of cancer. The fact thatthe sera of all seven patients with advanced cancers of the lipand oropharynx, areas commonly infected by HSV 1, reactedonly with HSV 1 and not at all with HSV 2 nonvirion antigens(Table 3) is remarkably in accord with the observations thathigh-multiplicity infection of tissue culture cells with HSV 1strains has induced antigens which on hyperimmunization ofguinea pigs produced antibodies that react with HSV 1 andnot with HSV 2 nonvirion antigens (13). These results sug-gest the possibility of a special involvement of HSV 1 in thesecancers, unlike that associated with ordinary primary andrecurrent infections with this virus. Similarly, the fact thatthe sera of all 14 patients with advanced carcinoma of thecervix uteri or vulva, areas commonly infected by HSV 2,reacted with both HSV 1 and HSV 2 nonvirion antigens isremarkably in accord with the observations that high-multi-

during virion absorption. A described procedure for the com-

Proc. Nat. Acad. Sci. USA 70 (1978)

plicity infection of tissue cultures with HSV 2, which has in-

Dow

nloa

ded

by g

uest

on

Feb

ruar

y 18

, 202

0

3228 Medical Sciences: Sabin and Tarro

TABLE 2. Complement fixation tests with HSV 1 and HSV 2nonvirion antigens and sera of patients with advanced cancer in

different parts of the body

All in this group All in this group

positive* negative*

No. No.Site of cancer tested Site of cancer tested

Lip 2 Gum 1Mouth 1 Tongue 2Oropharynx 5 Tonsil 2Nasopharynx 10 Salivary gland 2Kidney 8 Accessory sinus 3Bladder 8 Epiglottis 1Prostate 8 Lung and bronchus 7Cervix uteri 13 Stomach 5Vulva 1 Colon 9

Breast 5Corpus uteri 5Ovary 5Testis 3Liver 4Thyroid 4Kidney-embryonal, Wilms' 4Melanoma 4Hodgkin's lymphoma 5Acute lymphocytic leukemia 5Acute myelocytic leukemia 5

Total patients 56 Total patients 81

* With HSV 1 only or with HSV 1 and HSV 2.

duced only HSV 2 nonvirion antigens demonstrable by com-

plement fixation has, nevertheless, on hyperimmunization ofguinea pigs produced antibodies for both HSV 2 and HSV 1nonvirion antigens (13). In these cancers, the results suggestthe possibility of a special involvement of HSV 2, unlikethat associated with ordinary primary and recurrent infectionswith this virus. The words "suggest the possibility" are usedhere instead of "indicate" to emphasize the need for estab-lishing, by appropriate tests, that the nonvirion antibodiesin these patients are actually in response to neoantigens codedby HSV genetic information and not equally well by any otherDNA virus or in response to antigens, coded by the host'sDNA, which may be repressed except during embryonicdevelopment and differentiation, e.g., the different varietiesof so-called carcinoembryonic antigens.

TABLE 3. Complement fixation reactivity of positive cancer

sera with HSV 1 and HSV 2 nonvirion antigens

No. positive with

Site of cancer No. tested HSV 1 HSV 2

Cervix 13 13 13Vulva 1 1 1Prostate 8 8 7 + 1?Bladder 8 8 6+ 1?Kidney 8 8 6 + 1?Nasopharynx 10 10 9 + 1?Oropharynx 5 5 0Mouth 1 1 0?

TABLE 4. Complementfixation antibody titers of positive humancancer sera with HSV 1 and HSV 2 nonvirion antigens

CF antibody titers* of individual sera in

Site No. corresponding order withof cancer tested HSV 1 HSV 2

Cervix uteri 13 32, 6, 4,4 4, 4, 4 32, 16, 8, 8, 8, 6, 44,4,4,4,4,4 4,4,4,4,4,4

Vulva 1 4 8Prostate 8 8, 6, 44, 4,4,4, 4 8, 4, 4,4,4 4,4, 0?Bladder 8 12, 88, 6, 6, 4, 4, 4 16, 4, 0, 0?, 4, 4, 8, 8Kidney 8 6, 4, 4, 4, 4, 4,4,4 8,88,8, 16, 4,0? 0Nasopharynx 10 8, 8, 8, 8, 8 6, 6, 6, 6, 4

4,4,4,4,44 4,4,4,4,0?Oropharynx 5 6, 4, 4, 4, 4 ,0, 0, 0, 0Mouth 1 4 0?Lip 2 4,4 0, 0

* Titer = dilution at which there was no hemolysis or less than50% of erythrocytes were hemolyzed. Titers of 6 or 12 signify 50%hemolysis at dilutions of 1:8 or 1:16 and no hemolysis or lessthan 50% hemolysis at 1:4 and 1:8, respectively. 0? = about50% hemolysis at 1:4 dilution with complete hemolysis in allcontrols. 0 = complete hemolysis (usually) or occasionally in-complete hemolysis with a residue of less than 33% of erythro-cytes at 1:4 dilution.

The following control tests were done to establish the speci-ficity of the antibodies in the positive cancer sera for non-virion antigens coded by HSV genetic material: (a) 10 Posi-tive sera (two each of nasopharynx, bladder, kidney, prostrate,and cervix uteri) failed to react with cells, harvested 10 minand 6 hr after high-multiplicity infection with the DNA vac-cinia virus, which produces, within 10 min (12), a transitorymarked rise in concentration of a preexisting normal antigenthat is also present in greater concentration in almost alltransformed and malignant human cells; (b) the antibodies inthe same 10 sera were not removed or diminished in titer bymassive absorption with freshly harvested HEp2 cells; and (c)the antibodies in four positive sera (two each of kidney andcervix uteri) were not removed or diminished in titer by mas-sive absorption with freshly trypsinized cells from the kidneysof 12- to 14-week-old human embryos. These results, takentogether with the other manifestations of specificity (Tables2 and 3), strongly support the conclusion that the antibodieswe demonstrated in the positive human cancer sera were in-deed against neoantigens coded in one group of cancers (oro-pharynx and lip) by HSV 1 and in the other group of cancers(cervix uteri, vulva, and with a few possible exceptions alsoprostate, bladder, kidney, and nasopharynx) by HSV 2. It is,therefore, all the more intriguing why all five oropharyngealcancers are so strictly associated only with HSV 1 while atleast nine of ten (and perhaps all) nasopharyngeal cancers areso definitely associated with HSV 2.The results in Table 4, showing the titers of the HSV 1 and

HSV 2 nonvirion antibodies for each positive serum, are ofinterest especially because of the large number of sera thatwere positive only at the 1:4 dilution. Had we started ourtests with a 1:10 dilution of serum and used the micro com-plement fixation procedure, as was done in the negative testswith various adenovirus T (tumor) antigens and unabsorbedsera (15) from the same collection that we used, 59%/ of our 56

Lip 2 2 0 positive sera with titers of 1: 6 or less and perhaps also 32% of

Proc. Nat. Acad. Sci. USA 70 (1973)

Dow

nloa

ded

by g

uest

on

Feb

ruar

y 18

, 202

0

Herpesviruses 1 and 2 and Human Cancers 3229

the sera with titers of 1:8 would have yielded negative resultsin our macro complement fixation test, in which four-timesmore serum is used than in the micro test. Low antibody re-

sponses by tumor-bearing animals to nonvirion antigens are

not unusual with some of the experimentally produced DNAvirus malignancies (9). The negative results that we ob-

tained with the sera of four women with early, limited, malig-nant changes in the cervix uteri, in contradistinction to thepositive results on all 13 women with advanced cancer of thecervix uteri, is an indication that the HSV nonvirion anti-bodies appear after extensive and perhaps continuous anti-genic stimulation. Huebner (16) reported the disappearanceof nonvirion antibodies after surgical excision of experimentaladenovirus-induced tumors in hamsters and their reap-

pearance after regrowth of the tumor.

DISCUSSION

The results presented here provide strong evidence for an

etiologic role of HSV 1 and HSV 2 in at least nine varieties ofhuman cancer. Relevant data in support of this conclusion are

contained in a recent publication (17), which reported that a

human carcinoma of the cervix uteri that had no infectiousvirus contained HSV messenger RNA and also a fragmentof HSV DNA linked to host DNA. Moreover, it was shownthat "about half of the viral DNA template transcribed inthe cervical tumor belongs to the set of sequences that are

shared in common between HSV 1 and HSV 2." Another re-

cent publication (18) reported positive complement fixationreactions of nonvirion antigens in HEp2 cells infected with an

HSV 2 strain and also of a special fraction from human car-

cinoma of the cervix uteri with our HSV 1 guinea pig 71serum which did not react with nonvirion antigens in cellsinfected with three HSV 2 strains (13); then these-reportedtests cannot be properly evaluated. The procedures reportedin the present and preceding communications (11-13) can now

be used for more extensive investigations of the role of HSVand other DNA viruses in various human cancers.

We thank Miss Louise Malan, Mr. Willis Foster, and duringthe latter part of the program Dr. Ann Boyd for their dedicatedhelp with the tremendous amount of laboratory work. Specialthanks are due to Dr. John Landon, Director for Science, Fred-erick Cancer Research Center operated by Litton Bionetics, Inc.for his help in the rapid mobilization of laboratory facilities,equipment, and manpower.

1. Huebner, R. J., Rowe, W. P., Turner, H. C. & Lane, W. T.(1963) Proc. Nat. Acad. Sci. USA 50, 379-389.

2. Sabin, A. B. & Koch, M. A. (1963) Proc. Nat. Acad. Sci.USA 50, 407-417.

3. Sabin, A. B. & Koch, M. A. (1964) Proc. Nat. Acad. Sci.USA 52, 1131-1138.

4. Rapp, F., Kitahara, T., Butel, J. S. & Melnick, J. L. (1964)Proc. Nat. Acad. Sci. USA 52, 1138-1142.

5. Pope, J. H. & Rowe, W. P. (1964) J. Exp. Med. 120, 577-588.

6. Hoggan, M. D., Rowe, W. P., Black, P. H. & Huebner, R. J.(1965) Proc. Nat. Acad. Sci. USA 53, 12-19.

7. Rapp, F., Butel, J. S., Feldman, L. A., Kitahara, T. &Melnick, J. L. (1965) J. Exp. Med. 121, 935-944.

8. Sabin, A. B. (1967) Memoires de L'Academie Royale deMedicine de Belgiques, Ie Sgrie 6, 61-81.

9. Sabin, A. B. (1967) in Specific Tumor Antigens, UICC Mono-graph Series, ed. Harris, R. J. C. (Munksgaard, Copen-hagen), Vol. 2, pp. 251-264, 328-332.

10. Sabin, A. B. (1968) Cancer Res. 28, 1849-1858.11. Tarro, G. & Sabin, A. B. (1970) Proc. Nat. Acad. Sci. USA

65, 753-760.12. Tarro, G. & Sabin, A. B. (1970) Proc. Nat. Acad. Sci. USA

67, 731-737.13. Tarro, G. & Sabin, A. B. (1973) Proc. Acad. Sci. USA 70,

1032-1036.14. Roizman, B. (1966) in Perspectives in Virology, ed. Pollard,

M. (Harper & Row, New York, N. Y.), Vol. 4, pp. 283-304.15. Gilden, R. V., Kern, J., Lee, Y. K., Rapp, F., Melnick,

J. L., Riggs, J. L., Lennette, E. H., Zbar, B., Rapp, H. J.,Turner, H. C. & Huebner, R. J. (1970) Amer. J. Epidemiol.91, 500-509.

16. Huebner, R. J. (1967) in Perspectives in Virology, ed. Pol-lard, M. (Academic Press, New York), Vol. 5, pp. 147-166.

17. Frenkel, N., Roizman, B., Cassai, E. & Nahmias, A. (1972)Proc. Nat. Acad. Sci. USA 69, 3784-3789.

18. Hollinshead, A. C. & Tarro, G. (1973) Science 179, 698-700.

Proc. Nat. Acad. Sci. USA 70 (1973)

Dow

nloa

ded

by g

uest

on

Feb

ruar

y 18

, 202

0