HSV presence in brains of individuals without dementia : the TASTY brain series

RESEARCH ARTICLE HSV p esence in b ains o indi iduals wi hou demen ia: he TASTY b ain se ies Jan Olsson 1 , Hugo Lo  heim 2 , Emma Honkala 1 , Pekka J. Ka hunen 3 , F ed ik Elgh 1 and Eloise H. Kok 3, * ABSTRACT He pes simplex i us (HSV) ype 1 a ec s a majo i y o he popula ion and ecen e idence sugges s in ol emen in Alzheime ’s disease ae iology. We in es iga ed he p e alence o HSV ype 1 and 2 in he Tampe e Au opsy S udy (TASTY) b ain samples using PCR and se o-posi i i y in plasma, and associa ions wi h Alzheime ’s disease neu opa hology. HSV was shown o be p esen in human b ain issue in 11/584 (1.9%) o samples in he TASTY coho , o which six had Alzheime ’s disease neu opa hological amyloid be a (Aβ) agg ega ions. Addi ionally, se ological da a e ealed 86% o se um samples es ed we e IgG-posi i e o HSV. In conclusion, we epo epidemiological e idence o he p esence o HSV in b ain issue ee om encephali is symp oms in a coho mos closely ep esen ing he gene al popula ion (a minimum p e alence o 1.9%). Whe eas 6/11 samples wi h HSV DNA in he b ain issue had Aβagg ega ions, mos o hose wi h Aβagg ega ions did no ha e HSV p esen in he b ain issue. KEY WORDS: He pes simplex i us, Amyloid be a agg ega ions, Alzheime ’s disease, PCR de ec ion, Human b ain issue, Pa a in-embedded samples INTRODUCTION He pes simplex i us ype 1 (HSV1) a ec s a majo i y o he popula ion, in some coho s up o 90% (Lö heim e al., 2015b; Smi h and Robinson, 2002). The disease appea s as cold so es, ypically seen on he lips o ace, wi h p ima y in ec ion usually du ing childhood. HSV ype 2 (HSV2) p edominan ly a ec s he geni al a ea and is one o he mos common sexually ansmi ed in ec ions. The i us emains la en in neu onal cell bodies and eac i a es due o s ess, illness and o he unknown ac o s h oughou an indi idual’s li e. In some cases, indi iduals can de elop ad e se eac ions such as he pes simplex encephali is (HSE), a li e- h ea ening disease (Whi ley and Roizman, 2001). Alzheime ’s disease (AD) is he mos common o m o demen ia and g owing elde ly popula ions will con inue o pu p essu e on heal h sys ems o alle ia e his de as a ing disease, which cu en ly has no cu e (Jellinge , 2006). The disease is cha ac e ised by p og essi e demen ia, culmina ing in neu onal loss hough o be caused by he wo main hallma ks o he disease –amyloid be a (Aβ) agg ega ions and neu o ib illa y angles (NFT) (Goa e and Ha dy, 2012). The mos common and s onges gene ic isk ac o o AD is he apolipop o ein E (APOE) epsilon 4 allele, which inc eases he isk o de eloping AD app oxima ely h ee old in he e ozygo es and e en mo e among homozygo es (Co de e al., 1995; Fa e e al., 1997; an Duijn e al., 1994). This gene can also acili a e in ec ion wi h HSV (Bha acha jee e al., 2008; Bu gos e al., 2006, 2003). In addi ion, he a e HSE a ec s simila b ain egions o hose seen in AD (Ball, 1982). In ecen yea s, e idence o HSV as a possible cause o AD has accumula ed, and is suppo ed by epidemiological (Le enneu e al., 2008; Lö heim e al., 2015a,b), neu opa hological (S eel and Eslick, 2015; Wozniak e al., 2009), as well as in i o da a demons a ing AD-speci ic changes in neu al cells ollowing HSV in ec ion (San ana e al., 2012). We in es iga ed he p e alence o HSV in 584 b ain issue samples om Tampe e Au opsy S udy (TASTY) coho o 603 indi iduals, p ima ily wi hou demen ia, o assess he i us’abili y o gain access o he b ain wi hou causing no iceable symp oms such as HSE. In addi ion, we in es iga ed whe he he p esence o HSV in he b ain had any e ec on he p esence o he neu opa hology associa ed wi h AD in his coho o communi y-dwelling indi iduals p ima ily ee om demen ia. RESULTS PCR de ec ion o HSV b ain issue posi i i y DNA was ex ac ed om 584 pa a in-embedded b ain samples om he TASTY coho , comp ising pooled middle on al gy us, gy us cinguli wi h co pus callosum, hippocampus and ce ebellum, o es o he p esence o HSV genomic ma e ial. To al amoun o ex ac ed DNA o indi idual samples, as assessed by spec opho ome y, anged om 1.8-57.0 µg. As he spec opho ome ic alue does no di e en ia e be ween in ac and deg aded DNA, we employed a qPCR-based assay designed o de e mine, in absolu e numbe s, each sample’s con en o a 41 bp a ge wi hin a conse ed single- copy locus in he human genome. All samples we e quan i ied agains he 41 bp a ge eac ion and numbe s o gene copies/sample we e ound o ange om 1.2×10 5 -48.6×10 5 . Samples had simila human gene copies/ eac ion whe he posi i e o he p esence o HSV (n=11, con aining 1.2×10 5 -20.2×10 5 copies; mean 10.6×10 5 ), o nega i e (n=573, con aining 1.2×10 5 -48.5×10 5 copies; mean 14.7×10 5 )(P=0.127).All HSV1- o HSV2-posi i e samples we e also quan i ied by means o a 129 bp a ge wi hin he same conse ed single-copy locus in he human genome, and calcula ing he a io o 129/41 bp gene copies allowed o u he assessmen o DNA quali y; as DNA agmen a ion des oys longe empla es a a highe a e han sho e ones, his alue declines as DNA quali y dec eases. The 11 HSV-posi i e samples sco ed 4.5-21.0% (mean 12.6%) compa ed wi h a ange o 2.4-15.5% (mean 7.8%) o 24 andomly selec ed nega i e samples (P=0.039). Recei ed 21 June 2016; Accep ed 20 Sep embe 2016 1 Depa men o Clinical Mic obiology, Vi ology, Umeå Uni e si y, Umeå 90185, Sweden. 2 Depa men o Communi y Medicine and Rehabili a ion, Ge ia ic Medicine, Umeå Uni e si y, Umeå 90185, Sweden. 3 Depa men o Fo ensic Medicine, Uni e si y o Tampe e, Tampe e 33520, Finland. *Au ho o co espondence ([email p o ec ed]) E.H.K., 0000-0001-7399-9749 This is an Open Access a icle dis ibu ed unde he e ms o he C ea i e Commons A ibu ion License (h p://c ea i ecommons.o g/licenses/by/3.0), which pe mi s un es ic ed use, dis ibu ion and ep oduc ion in any medium p o ided ha he o iginal wo k is p ope ly a ibu ed. 1349 © 2016. Published by The Company o Biologis s L d | Disease Models & Mechanisms (2016) 9, 1349-1355 doi:10.1242/dmm.026674 Disease Models & Mechanisms PCR eac ions di ec ed agains 62-73 bp sec ions o he HSV1 and HSV2 genomes we e used o de ec he espec i e i al DNA in he b ain issue samples. HSV1 PCR eac ions a ge ed he US5,UL5, and UL27 genes; HSV2 PCR eac ions a ge ed US4, UL5 and UL29. In he ade-o be ween maximal sensi i i y on he one hand and high s anda ds o speci ici y and ep oducibili y on he o he , no C cu -o alue was employed, and only samples posi i e o a leas wo o he PCR eac ions we e sco ed posi i e. The independen PCR eac ions he eby con i m each o he (Mahony e al., 1993). O 584 analysed samples, 10 we e posi i e o HSV1, whe eas one was posi i e o HSV2 (posi i i y in US4 and UL29, bu no UL5 PCR eac ions). None we e posi i e o bo h HSV1 and HSV2. In de ec ing HSV1 DNA, he US5,UL5 and UL27 PCR eac ions we e posi i e in 18 (3.1%), 16 (2.7%) and 19 (3.3%) ou o 584 analysed cases, espec i ely. The e we e 27 cases (4.6%) wi h only one posi i e eac ion, and 10 (1.7%) whe e a leas wo eac ions we e posi i e. Six ou o hese 10 we e iple-posi i e. Mean C alues we e 38.0±2.7 (±s.d.) o cases wi h a leas wo posi i e eac ions compa ed wi h 39.4±1.0 o hose wi h only one posi i e eac ion (P=0.156). These C alues a e high, indica ing sub- op imal qPCR condi ions, p obably due o he p esence o inhibi o y subs ances. Howe e , aga ose-gel elec opho esis s aining showed unambiguous co ela ion be ween a posi i e qPCR sco e and he p esence o a p oduc band o co ec size (da a no shown). The co ela ions be ween posi i i y o he di e en PCR eac ions we e 0.395 (US5 o UL5), 0.383 (UL5 o UL27) and 0.414 (UL5 o UL27), espec i ely (P<0.001 o all co ela ions). In he ollowing analyses, he HSV1 and HSV2 PCR da a a e pooled (see Table 1 o HSV DNA-posi i e case cha ac e is ics). HSV b ain issue posi i i y Cases (n=10 HSV1 and 1 HSV2, o al o n=11 HSV-posi i e) who had a leas wo posi i e PCR eac ions we e ega ded as HSV DNA-posi i e in b ain issue (11/584=1.9%). Age, sex and APOEε4ca ie ship (P=0.413, P=0.754 and P=0.746, espec i ely) did no a ec he isk o being HSV DNA-posi i e. Indi iduals we e aged 60 yea s and o e in 59% o he cases (n=346), o which 2.0% (7/346) had a p e alence o HSV posi i i y in b ain issue. The p opo ion o cases wi h Aβagg ega ions (Fig. 1) was 6/11 (54.5%) among hose posi i e o HSV DNA in b ain issue compa ed wi h 160/530 (30.2%) among hose nega i e o HSV DNA (Fishe ’s exac es , P=0.101 o cases wi h a ailable Aβagg ega ion da a), and o NFT, 5/9 (55.6%) compa ed wi h 194/ 465 (41.7%) (Fishe ’s exac es , P=0.502 o cases wi h NFT da a). The p opo ion o cases wi h bo h Aβagg ega ions and NFT was 4/9 (44.4%) among HSV DNA-posi i e cases compa ed wi h 81/451 (18.0%) (Fishe ’s exac es , P=0.097 o cases wi h comple e Aβ agg ega ion and NFT neu opa hology da a). Six cases o he TASTY coho had been diagnosed wi h AD while ali e, o which one indi idual was posi i e o HSV DNA in b ain issue. The p opo ion o indi iduals wi h HSV DNA posi i i y was 1/6 (16.7%) among hose wi h AD, compa ed wi h 10/551 (1.8%) o non-demen ed cases (Fishe ’s exac es , P=0.113 o cases wi h documen ed demen ia in o ma ion). Se ology I was possible o ob ain se ology da a o 141 o he cases (24.1%), o which 121 (85.8%) we e posi i e o an i-HSV IgG. Mean age was 57.6±15.3 o an i-HSV-IgG-nega i e cases and 64.5±17.5 o hose an i-HSV-IgG-posi i e (P=0.103). Sex and APOEε4 ca ie ship did no impac on he isk o ca ying HSV; 74/88 (84.1%) o he men and 47/53 (88.7%) o he women we e an i- HSV-posi i e (P=0.449), and 45/53 (84.9%) o he APOEε4 ca ie s compa ed wi h 76/88 (86.4%) o hose no ca ying an APOEε4allele (P=0.810). Se ology and neu opa hology Among an i-HSV-IgG-posi i e cases o which Aβagg ega ion da a was a ailable, 36/111 (32.4%) we e Aβ-agg ega ion-posi i e compa ed wi h 2/19 (10.5%) o hose an i-HSV-IgG-nega i e [χ 2 : P=0.052; adjus ed o age in a bina y logis ic eg ession: odds a io (OR) 2.771, 95% con idence in e al (CI) 0.567–13.541, P=0.208]. NFT was seen among 30/77 (39.0% o an i-HSV-IgG-posi i e) compa ed wi h 3/12 (25.0% o an i-HSV-IgG-nega i e) wi h NFT neu opa hology da a in his coho (χ 2 :P=0.352; age-adjus ed: OR 1.264, CI 0.281–5.681, P=0.760). IgM and IgG se ology The e we e 15/141 (2.5%) an i-HSV-IgM-posi i e. All bu one o hem we e also an i-HSV-IgG-posi i e and hence ega ded as eac i a ed in ec ions. One 58-yea -old male was IgM-posi i e and IgG-nega i e, which indica es p ima y in ec ion. The indi idual had no Aβagg ega ions o NFT. Among hose posi i e o an i-HSV IgG, hose who we e also posi i e o an i-HSV IgM we e younge (55.3±16.7 e sus 65.6±17.3, P=0.038) and mo e o en male [12/74 (16.2%) e sus 2/47 (4.3%), P=0.045]. An i-HSV IgM an ibodies we e equally common among hose wi h o wi hou an APOEε4 Table 1. Cases om he Tampe e Au opsy S udy ha we e posi i e o HSV DNA Case # Age Sex APOE Aβ% Plaque sub ype CERAD* NFT/1 mm 2 Demen ia HSV1 posi i e ‡ HSV2 posi i e ‡ HSV IgG HSV IgM CoD_class 165 43 Male e3/e3 0 None No AβND None 1 0 Neg Neg Suicide 174 82 Male e3/e3 0 None No AβND None 1 0 92 Neg Disease (cance ) 226 78 Female e4/e3 1.1 ND Mode a e Aβ5.6 AD 0 1 Neg Neg Acciden 270 74 Male e3/e3 0.8 Classic F equen Aβ4.8 None 1 0 ND ND Disease (hea disease) 411 72 Male e4/e4 0 None No Aβ0 None 1 0 ND ND Disease (in ec ious, no HSE) 415 50 Male e3/e3 0 None No Aβ3.2 None 1 0 ND ND Disease (hea disease) 452 89 Female e3/e3 0 None No Aβ0 None 1 0 ND ND Acciden 459 78 Female e3/e2 0.4 P imi i e Spa se Aβ0 None 1 0 ND ND Disease (s oke) 504 59 Male e3/e4 1 Classic Mode a e Aβ13.6 None 1 0 ND ND Disease (hea disease) 513 57 Male e3/e3 0.3 P imi i e Spa se Aβ0 None 1 0 ND ND Disease (in ec ious, no HSE) 561 68 Male e3/e4 1.3 Classic Mode a e Aβ14.4 None 1 0 ND ND Disease (hea disease) *Densi y o Aβagg ega ions ‡ Posi i e esul om wo HSV PCR eac ions AD, Alzheime ’s disease; CERAD, Conso ium o Es ablish A Regis y o Alzheime ’s Disease, CoD, cause o dea h; HSE, He pes simplex encephali is; ND, no da a; Neg, nega i e 1350 RESEARCH ARTICLE Disease Models & Mechanisms (2016) 9, 1349-1355 doi:10.1242/dmm.026674 Disease Models & Mechanisms allele [6/45 (13.3%) e sus 8/76 (10.5%), P=0.641]. Adjus ed o age and sex, he e we e no signi ican ela ions be ween an i-HSV IgM and Aβagg ega ions o NFT (P=0.346 and P=0.519, espec i ely). Se ological le els and co ela ions An i-HSV IgG le el (among hose an i-HSV-IgG-posi i e) was a ec ed by age and sex o he cases (mul iple linea eg ession: age β=0.298, P=0.018, male sex β=11.563, P=0.011). The le el was 81.8±23.0 among hose who ca ied a leas one APOEε4allele compa ed wi h 74.2±22.6 among hose who did no (P=0.079). The e we e no signi ican co ela ions be ween an i-HSV IgG le el and Aβ-immuno eac i i y (IR) pe cen age and NFT coun (Pea son co ela ion −0.002, P=0.987 and –0.039, P=0.734, espec i ely). Se ology and PCR posi i i y Blood plasma was a ailable o only h ee o he HSV-DNA- posi i e samples. In e es ingly, wo ou o hese h ee cases –one HSV1-posi i e and he single HSV2-posi i e –we e nega i e o an i-HSV IgG, compa ed wi h 120/138 (87.0%) o hose HSV- DNA-nega i e wi h a ailable se ological da a (Fishe ’s exac es P=0.053). DISCUSSION Ou main inding was ha in an unhospi alised coho (see Fig. 2) o 603 indi iduals p ima ily wi hou demen ia, we de ec ed HSV in b ain issue o 11 (1.9%) o he 584 cases om which i was possible o ex ac DNA. Ou se ological analyses o 141 cases e ealed 85.8% ca ying HSV, which is sligh ly lowe han –al hough in line wi h –p e ious epidemiological s udies o he i us’p esence (Lö heim e al., 2015a,b; Smi h and Robinson, 2002). We used s ic c i e ia o posi i i y, possibly unde es ima ing he ac ual numbe because o sample deg ada ion. P epa a ions om FFPE b ain issue gene ally yield mo e se e ely deg aded DNA han p epa a ions om, o example, esh ozen issue (Fe e e al., 2007). In addi ion, o malin s o age leng h has been sugges ed o a ec genomic DNA yield (Wang e al., 2013; Fe e e al., 2007), in which he cu en samples we e ixed o a ound 2 weeks (Kok e al., 2009). We assessed he quali y o p epa ed human DNA as an indica o o chances o success ul HSV DNA de ec ion, assuming ha DNA deg ada ion a ec s human and i al DNA equally. In ac , samples posi i e o HSV DNA had a sligh ly highe a e age 129/41 bp human DNA a io han hose nega i e o HSV DNA, possibly indica ing ha some samples a e alse nega i es. The indica ed 1.9% o HSV-posi i e samples should he e o e be ega ded as a minimum es ima e, especially in ligh o ea lie s udies ha ha e ocussed on AD pa ien s in compa ison wi h non-AD con ols. Such s udies ha e epo ed a HSV p e alence anging om 21.7-100% o non-AD indi iduals (F ase e al., 1981; Be and e al., 1993; Ba inge and Pisani, 1994; I zhaki e al., 1997; Jamieson e al., 1991, 1992; Wozniak e al., 2005), al hough hey ha e gene ally deal wi h olde indi iduals and subs an ially smalle sample sizes. Sequence a ia ions o he in ec ing HSV s ains migh also ha e esul ed in missed in ec ions. I should also be no ed ha he coho s ems om a Finnish popula ion no p e iously explo ed o HSV DNA. To he au ho s’knowledge, a s udy o his size has no p e iously been pe o med, and i sheds ligh on he p e alence o HSV in b ain issue in a unique coho mos closely ep esen ing he gene al popula ion, including indi iduals o all ages. The esul s sugges HSV p esence in b ain issue is no unusual, and u he indica es he exis ence o pe sis en o la en in ec ions wi hin he cen al ne ous sys em (CNS) wi hou p oducing encephali is symp oms. The esul s also indica e ha sp ead o he CNS is much mo e common o HSV1 han o HSV2, consis en wi h a low p opo ion o HSV2 being ound in human b ain in o he s udies (Lin e al., 2002). The small numbe o HSV-DNA-posi i e samples esul ed in a lack o powe in mos analyses. Al hough in e p e a ion mus be made wi h app op ia e cau ion, some non-signi ican esul s migh s ill be wo h commen ing on. The p opo ion ha ing Aβ agg ega ions among hose who had HSV DNA in b ain issue (i.e. es ed posi i e in wo o mo e PCR analyses) we e 6/11 (54.5%) compa ed wi h 30.2% among hose who we e HSV-DNA-nega i e (P=0.101), 16.7% o hose wi h AD had HSV DNA p esen in b ain issue compa ed wi h 1.8% o non-demen ed cases (P=0.113), and an i-HSV se o-posi i i y was close o being signi ican ly associa ed wi h an inc eased isk o Aβagg ega ions (P=0.052). These esul s a e no in disag eemen wi h p e ious s udies sugges ing ha Aβ agg ega ions and/o AD a e associa ed wi h HSV (Ball, 1982; I zhaki, 2014; I zhaki e al., 2016; Le enneu e al., 2008; Lö heim e al., 2015a,b; S eel and Eslick, 2015), bu should be con i med in la ge s udies. Howe e , mos o he indi iduals ha had Aβ agg ega ions did no ha e de ec able HSV DNA in b ain issue. Al hough his migh in pa be explained by he sensi i i y o ou analyses, ecen s udies a e poin ing owa ds he Aβ eac ion being a pa o he inna e immune sys em (I zhaki e al., 2016; Kuma e al., 2016), wi h a b oad an imic obial e ec agains bac e ia, ungi and 4 % 96 % Aβ + e 1 % 99 % Aβ - e HSV DNA + e HSV DNA - e HSV DNA + e HSV DNA - e Fig. 1. The p e alence o HSV DNA posi i i y in cases sepa a ed by Aβ pa hology p esence. Cases wi h HSV DNA p esen a e depic ed in da k g ey, e sus hose wi hou HSV DNA, spli in o g oups wi h and wi hou Aβposi i i y. 1351 RESEARCH ARTICLE Disease Models & Mechanisms (2016) 9, 1349-1355 doi:10.1242/dmm.026674 Disease Models & Mechanisms i uses including HSV (Bou gade e al., 2015; Soscia e al., 2010; Whi e e al., 2014). Se e al bac e ia and i uses ha e been demons a ed as being able o igge he Aβ eac ion in cul u ed cells and mice (K is en e al., 2015; Li le e al., 2014; Miklossy e al., 2006; San ana e al., 2012; Shipley e al., 2005). A possible explana ion migh he e o e be ha p e ious in ec ions lea e sca s in he o m o Aβagg ega ions; howe e , ce ain pe sis en in ec ions including HSV migh con inue o igge he Aβ p oduc ion o p olonged pe iods o ime o p oduce accumula ion o Aβand ul ima ely Alzheime ’s disease. A numbe o ecen s udies ha e lean weigh o suppo he heo y o a pa hogen- media ed AD hypo hesis (Bou gade e al., 2016; Miklossy, 2011). These include in es iga ions in o ungi (Pisa e al., 2015), bac e ia (Miklossy, 2015; Po gie e e al., 2015) and e en hei by-p oduc s (Kell and P e o ius, 2015), and poin owa ds a po en ial ole o se e al in ec ious pa hogens in AD neu odegene a ion. Two ou o h ee indi iduals posi i e o HSV DNA in b ain issue wi h se a a ailable we e HSV se o-nega i e, a bo de line signi ican ly lowe p opo ion (P=0.053) compa ed wi h hose nega i e o HSV DNA, among whom 87.0% we e HSV se o- posi i e. Al hough his unexpec ed esul mus be in e p e ed wi h cau ion, one possible explana ion could be indi iduals ha o some eason ha e no de eloped a p ope immune esponse o HSV in ec ion (he e p esen ed as low le els o speci ic an ibodies), migh ha e an inc eased isk o HSV sp ead o he CNS. Al e na i ely, sample deg ada ion and in e e ence wi h he ELISA sys em esul ing om pos -mo em-al e ed blood samples could be a sou ce o e o ; howe e , he high p e alence o HSV se o- posi i i y among hose nega i e o HSV DNA in b ain issue indica es a su icien sensi i i y in he ELISA analyses also o pos - mo em samples. In conclusion, we epo epidemiological e idence o HSV p esence in b ain issue om indi iduals who did no show encephali is symp oms in a coho mos closely ep esen ing he gene al popula ion (p e alence 1.9%). Six ou o 11 wi h HSV DNA in b ain issue had Aβagg ega ions, al hough mos o hose wi h Aβagg ega ions did no ha e HSV p esen in b ain issue. MATERIALS AND METHODS Coho and neu opa hology The TASTY coho has been desc ibed in de ail elsewhe e, including Aβ agg ega ion and neu o ib illa y angle (NFT) da a acquisi ion (Kok e al., 2009). B ie ly, 603 indi iduals (388 males, 215 emales; 64% male) aged 0-97 yea s (a e age 63 yea s) unde wen au opsy a he Depa men o Fo ensic Medicine, Uni e si y o Tampe e, Finland om 2002-2004 (see Table 2). Samples unde wen sec ioning and s aining wi h Bielschowsky sil e s ain acco ding o s anda dised p ocedu es, and we e measu ed as posi i e o Aβagg ega ions (n=541, 92.6% o he TASTY coho we e es ed; n=166, 30.7% we e posi i e) and Aβ-IR (immuno eac i i y) as a pe cen age o a ea co e ed by Aβagg ega ions (a e age Aβ-IR was 0.45%, ange 0-5.4%), and neu o ib illa y angle (NFT) da a (da a a ailable o n=474, 81.2%; n=199, 42.0% posi i e; NFT/1 mm 2 a e age=3.4, ange 0-60.8) (Kok e al., 2009). O he 603 cases in he TASTY coho , 31.1% we e APOEε4ca ie s (n=187), 584 had b ain issue blocks and 141 had se ology samples a ailable o he cu en s udy. No all cases had all da a a ailable owing o una ailable o missing samples, mos o en caused by cause o dea h hampe ing sample collec ion. 0 10 20 30 40 50 60 70 80 0-40 yea s 40.5 - 50 yea s 50.5 - 60 yea s 60.5 - 70 yea s 70.5 - 80 yea s 80.5+ yea s % a ec ed Age g oups No neu opa hology Aβ agg ega ions NFT Aβ & NFT AD HSV DNA posi i i y n=72 27.8% APOEε4 n=71 33.8% APOEε4 n=107 25.2% APOEε4 n=100 37% APOEε4 n=134 34.6% APOEε4 n=119 28% APOE ε4 Fig. 2. TASTY coho cha ac e is ics. Changes in age dis ibu ion o HSV DNA p esence in b ain issue, APOEε4ca ie ship, Aβagg ega ions, NFT, and combined neu opa hology, as well as Alzheime ’s disease cases o he TASTY coho . Please no e ha only hose cases wi h neu opa hology da a o bo h Aβ agg ega ions and NFT we e included in he igu e. 1352 RESEARCH ARTICLE Disease Models & Mechanisms (2016) 9, 1349-1355 doi:10.1242/dmm.026674 Disease Models & Mechanisms The au opsy se ies and i s use we e app o ed by he Na ional Au ho i y o Medicolegal A ai s in Finland (1239/32/200/01). The Regional E hical Re iew Boa d in Umeå, Sweden app o ed he s udy (2013/277-31). DNA ex ac ion Pa a in-embedded issue blocks o ou b ain egions (middle on al gy us, gy us cingula wi h co pus callosum, hippocampus and ce ebellum) we e sampled 4-6 imes wi h 20-µm- hick sec ions and ans e ed in o s e ile mic o uge ubes. Tissue was depa a inised acco ding o he guidelines ou lined by G ee e al. (1994). B ie ly, 1 ml xylene was added o each ube, cen i uged a 16,000 g o 5 min, a e which he supe na an was disca ded. This p ocedu e was ca ied ou wice. The xylene was hen emo ed in h ee washes wi h 99% e hanol wi h cen i uga ion a 16,000 g o 5 min in each washing s ep. Tubes we e le open in a dus -p o ec ed cabine o one hou o allow he emaining e hanol o e apo a e. The ea e , DNA was ex ac ed wi h a LIAISON Ix pipe ing/ex ac ion ins umen i ed wi h a LIAISON Ix DNA Ex ac ion ki (Diaso in). DNA P e ea men Bu e 2 supplemen ed wi h P o einase K (200 µl+10 µl was added o each sample, and he ubes we e incuba ed a 56°C o e nigh . Samples we e hen cen i uged (10 min, 20,000 g) o sedimen esidual ma e ial. DNA was p epa ed om he diges ed samples and elu ed in o 50 µl elu ion bu e . Quali y assessmen o ex ac ed DNA The o al amoun and pu i y o DNA o each sample was assessed by spec opho ome y (NanoD op 1000 Full Spec um UV/Vis Spec opho ome e , Wilming on, USA). The o al amoun o DNA was ob ained in ng/μl, and he A260/280 a io was calcula ed as an indica o o p o ein impu i ies. In o de o assess DNA quali y wi h espec o agmen a ion, samples we e subjec ed o a qPCR-based quan i ica ion o 41 bp and 129 bp a ge s wi hin a conse ed single-copy locus in he human genome using a KAPA Human Genomic DNA Quan i ica ion and QC Ki (KAPA Biosys ems, Cape Town, Sou h A ica). Samples we e dilu ed 1:100 in 10 mM T is pH 8.0+0.05% Tween 20 (DNA dilu ion bu e ) o all wi hin he dynamic ange o he assay, and he qPCR eac ion was pe o med acco ding o he manual o he p oduc on an ABI S epOne Plus ins umen . Absolu e quan i ica ion o 41 bp and 129 bp a ge s o human genomes in samples was achie ed using he s anda d p o ided in he ki . PCR me hods PCR eac ions di ec ed agains conse ed egions o he HSV1 and HSV2 genomes we e used o de ec he espec i e i al DNA in he samples. P ime s and TaqMan p obes we e designed wi h he online so wa e P ime 3Plus (Un e gasse e al., 2007). C ucial design pa ame e s we e: T m o p ime s 60°C, T m o p obes 70°C, leng h o ampli ied egions 60-75 bp, and no sequence ma ch o o he e e ence genomes. P ime s and p obes we e o de ed om Eu o insGenomics, Ebe sbe g, Ge many, wi h he sequence o p ime s and p obes, and he size o amplicons desc ibed in Table 3. DNA p epa ed om in-house HSV1 o HSV2 cul i a ions we e used as posi i e con ols. DNA was dilu ed 1:10, a e which 10 µl was added o a inal olume o 25 µl PCR mix con aining TAQMAN UNIV Mas e MIX (Li e Technologies, Ca lsbad, CA, USA), 0.3 µM o each p ime and 0.2 µM p obe. Ampli ica ion eac ions included an ini ial 15-min dena u a ion s ep a 95°C, ollowed by 43 cycles o 15 s a 94°C, and 60 s a 60°C. The eac ions we e pe o med on an ABI 7900HT ins umen , and esul s we e analysed wi h he SDS so wa e se o au oma ic baseline and h eshold (Au oma ic C ). Se ology Plasma collec ed a au opsy was s o ed a −80°C. Samples we e hawed, clea ed o pa icula ma e s by cen i uga ion (10 min, 20,000 g), and analysed o an i-HSV IgG and IgM an ibodies using ELISA as desc ibed p e iously (Lö heim e al., 2015a), wi h he excep ion ha a new HSV1 isola e (Umeå clinical isola e 3458-13) has been in oduced o an igen p oduc ion. The IgG an ibody ac i i y o he indi idual samples was exp essed in a bi a y uni s (AU), achie ed by di iding he sample’s abso bance wi h ha o a posi i e e e ence sample, and mul iplied by 100. Samples wi h IgG alues o 5 AU o abo e we e ega ded posi i e o HSV Table 2. Cha ac e is ics o he Tampe e Au opsy S udy (TASTY) coho indi iduals wi h b ain issue samples es ed o HSV DNA HSV-DNA-nega i e HSV-DNA-posi i e P- alue A e age age, yea s 62.6 (0-97 yea s) 68.1 (43-89 yea s) 0.413* B ain issue B ain, g ams 1408 (427-1910 g) 1364 (1042-1648 g) 0.620* Sex, male 367 (64.0%) 8 (72.7%) 0.754 (FE) BMI 27.4 (11.8-59.7) 26.9 (16.5-36.3) 0.899* Aβ-IR, % 0.44 (0-5.4%) 0.55 (0-1.3%) 0.057* NFT coun /mm 2 3.32 (0-60.8/mm 2 ) 4.62 (0-14.4/mm 2 ) 0.280* Aβagg ega ions 160 (30.2%) 6 (54.5%) 0.101 (FE) NFT 194 (41.7%) 5 (55.6%) 0.502 (FE) APOEε4ca ie s 177 (31.0%) 4 (36.4%) 0.746 (FE) Se ology HSV IgG ‡ 120 (87.0%) 1 (33.3%) 0.053 (FE) HSV IgM ‡ 15 (10.9%) 0 1.000 (FE) CoD 0.800 (FE) Disease 320 (56.0%) 8 (72.7%) Acciden 169 (29.6%) 2 (18.2%) Suicide 70 (12.3%) 1 (9.1%) Homicide 3 (0.5%) 0 Unknown 9 (1.6%) 0 BMI, body mass index; CoD, cause o dea h; IgG, immunoglobulin G se o- posi i i y o HSV; IgM, immunoglobulin M se o-posi i i y o HSV; FE, Fishe ’s exac es *Mann–Whi ney es ‡ This da a was only a ailable o a subse o cases o he TASTY coho Table 3. PCR eac ions used in his s udy agains HSV DNA Fo wa d p ime Re e se p ime P obe Amplicon leng h Re e ence HSV1 Reac ion 1 (US5) GGCCTGGCTATCCGGAGA GCGCAGAGACATCGCGA FAM–CAGCACACGACTTGGCGT TCTGTGT–TAMRA 63 bp Filen e al., 2004 Reac ion 2 (UL5) GCAGATGAGGTACGTGAGCA GACCTTCGAGCACCAGAAAC FAM–GTTCTCGCTCTGGCGGACGG AAC–TAMRA 70 bp This s udy Reac ion 3 (UL27) GCGCTGTATGTGGTTGTACG TCAAGACCACCTCCTCCATC FAM–TAAACTGCAGCCGGGCGAAC TC–TAMRA 62 bp This s udy HSV2 Reac ion 1 (US4) AGATATCCTCTTTATCATC AGCACCA TTGTGCTGCCAAGGCGA FAM–CAGACAAACGAACGCCGC CG–TAMRA 73 bp Filen e al., 2004 Reac ion 2 (UL5) AACCCAAACACCATCTTTCG TCACGTACGTCCTCAACAGC FAM–CGCGGTCACCGCGACC TG–TAMRA 64 bp This s udy Reac ion 3 (UL29) CACCAGCTGCTTGATGTTGT GGGAGGCTGGAGACGATTAT FAM–ACCACCGTGTGCAGGGCC TC–TAMRA 72 bp This s udy 1353 RESEARCH ARTICLE Disease Models & Mechanisms (2016) 9, 1349-1355 doi:10.1242/dmm.026674 Disease Models & Mechanisms IgG an ibody con en . IgM ELISA esul s we e dicho omously de e mined posi i e o nega i e wi h he cu -o alue se a ne abso bance ≥0.15. S a is ics SPSS o Windows ( e sion 23; IBM) was u ilised in s a is ical analyses. Va iables used we e Aβagg ega ions (p esen yes/no; Aβ-IR %), NFT (yes/ no, NFT coun ), APOEε4allele ca ie ship (yes/no), sex, HSV DNA posi i i y ( wo posi i e es s/no), an i-HSV IgG (p esen yes/no; AU), and an i-HSV IgM (p esen yes/no). Acknowledgemen s Thank you o Si kka Goebele o collec ion o au opsy samples comp ising he Tampe e Au opsy S udy (TASTY) coho . Compe ing in e es s The au ho s decla e no compe ing o inancial in e es s. Au ho con ibu ions P.J.K. collec ed he au opsy samples, E.H.K. p epa ed au opsy samples, pe o med s a is ical analyses and w o e he manusc ip , J.O. and E.H. pe o med PCR and ELISA expe imen s, H.L. pe o med s a is ical calcula ions, and F.E. p o ided me hodological insigh and ad ice o he expe imen al p ocedu es and manusc ip p epa a ion. All au ho s con ibu ed o he manusc ip and ha e app o ed he inal e sion. Funding This wo k was suppo ed by unding om he Suomen Kul uu i ahas o Pi kanmaan Rahas o (Finnish Cul u al Founda ion Pi kanmaa Regional Fund), The Company o Biologis s (Disease Models & Mechanisms), Va s e bo en La ns Lands ing (Va s e bo en Coun y Council), Kempes i else na (Kempe Founda ions), S e iges La ka o  bund ( he Swedish Medical Associa ion), he Swedish Demen ia Associa ion, T olle-Wach meis e Founda ion, he Demen ia Fund in Va s e bo en, he Swedish Alzheime Fund, Gun och Be il S ohnes S i else (S ohne Founda ion), Magnus Be g alls S i else (Be g all Founda ion) and Umeå Uni e si e Founda ion o Medical Resea ch. These unding sou ces had no in ol emen in he s udy design; collec ion, analysis and in e p e a ion o da a; w i ing o he epo ; o he decision o submi his a icle o publica ion. Re e ences Ball, M. J. (1982). Limbic p edilec ion in Alzheime demen ia: is eac i a ed he pes i us in ol ed? Can. J. Neu ol. Sci. 9, 303-306. Ba inge , J. R. and Pisani, P. (1994). He pes simplex i us genomes in human ne ous sys em issue analyzed by polyme ase chain eac ion. Ann. Neu ol. 36, 823-829. Be and, P., Guillaume, D., Hellaue , L., Dea, D., Lindsay, J., Kogan, S., Gau hie , S. and Poi ie , J. (1993). Dis ibu ion o he pes simplex i us ype 1 DNA in selec ed a eas o no mal and Alzheime ’s disease b ains: a PCR s udy. Neu odegene a ion 2, 201-208. Bha acha jee, P. S., Neumann, D. M., Fos e , T. P., Bouhanik, S., Clemen , C., Vinay, D., Thompson, H. W. and Hill, J. M. (2008). E ec o human apolipop o ein E geno ype on he pa hogenesis o expe imen al ocula HSV-1. Exp. Eye Res. 87, 122-130. Bou gade, K., Ga neau, H., Gi oux, G., Le Page, A. Y., Boc i, C., Dupuis, G., F os , E. H. and Fu lo p, T.J . (2015). β-amyloid pep ides display p o ec i e ac i i y agains he human Alzheime ’s disease-associa ed he pes simplex i us-1. Bioge on ology 16, 85-98. Bou gade, K., Dupuis, G., F os , E. H., Fu lo p, T. (2016). An i- i al p ope ies o Amyloid-βpep ides. J. Alzheime s Dis. 54, 859-878. Bu gos, J. S., Rami ez, C., Sas e, I., Bullido, M. J. and Valdi ieso, F. (2003). ApoE4 is mo e e icien han E3 in b ain access by he pes simplex i us ype 1. Neu o epo 14, 1825-1827. Bu gos, J. S., Rami ez, C., Sas e, I. and Valdi ieso, F. (2006). E ec o apolipop o ein E on he ce eb al load o la en he pes simplex i us ype 1 DNA. J. Vi ol. 80, 5383-5387. Co de , E. H., Saunde s, A. M., S i ma e , W. J., Schmechel, D. E., Gaskell, P. C., J ., Rimmle , J. B., Locke, P. A., Conneally, P. M., Schmade , K. E., Tanzi, R. E. e al. (1995). Apolipop o ein E, su i al in Alzheime ’s disease pa ien s, and he compe ing isks o dea h and Alzheime ’s disease. Neu ology 45, 1323-1328. Fa e , L. A., Cupples, L. A., Haines, J. L., Hyman, B., Kukull, W. A., Mayeux, R., Mye s, R. H., Pe icak-Vance, M. A., Risch, N. and an Duijn, C. M. (1997). E ec s o age, sex, and e hnici y on he associa ion be ween apolipop o ein E geno ype and Alzheime disease. A me a-analysis. APOE and Alzheime Disease Me a Analysis Conso ium. JAMA 278, 1349-1356. Fe e , I., A ms ong, J., Capella i, S., Pa chi, P., A zbe ge , T., Bell, J., Budka, H., S o bel, T., Giaccone, G., Rossi, G. e al. (2007). E ec s o o malin ixa ion, pa a in embedding, and ime o s o age on DNA p ese a ion in b ain issue: a B ainNe Eu ope s udy. B ain Pa hol. 17, 297-303. Filen, F., S and, A., Alla d, A., Blombe g, J. and He mann, B. (2004). Duplex eal- ime polyme ase chain eac ion assay o de ec ion and quan i ica ion o he pes simplex i us ype 1 and he pes simplex i us ype 2 in geni al and cu aneous lesions. Sex. T ansm. Dis. 31, 331-336. F ase , N. W., Law ence, W. C., W oblewska, Z., Gilden, D. H. and Kop owski, H. (1981). He pes simplex ype 1 DNA in human b ain issue. P oc. Na l. Acad. Sci. USA 78, 6461-6465. Goa e, A. and Ha dy, J. (2012). Twen y yea s o Alzheime ’s disease-causing mu a ions. J. Neu ochem. 120, 3-8. G ee , C. E., Wheele , C. M. and Manos, M. M. (1994). Sample p epa a ion and PCR ampli ica ion om pa a in-embedded issues. Genome Res. 3, S113-S122. I zhaki, R. F. (2014). He pes simplex i us ype 1 and Alzheime ’s disease: inc easing e idence o a majo ole o he i us. F on . Aging Neu osci. 6, 529. I zhaki, R. F., Lin, W.-R., Shang, D., Wilcock, G. K., Fa aghe , B. and Jamieson, G. A. (1997). He pes simplex i us ype 1 in b ain and isk o Alzheime ’s disease. Lance 349, 241-244. I zhaki, R. F., La he, R., Balin, B. J., Ball, M. J., Bea e , E. L., B aak, H., Bullido, M. J., Ca e , C., Cle ici, M., Cosby, S. L. e al. (2016). Mic obes and Alzheime ’s disease. J. Alzheime s Dis. 51, 979-984. Jamieson, G. A., Mai land, N. J., Wilcock, G. K., C aske, J. and I zhaki, R. F. (1991). La en he pes simplex i us ype 1 in no mal and Alzheime ’s disease b ains. J. Med. Vi ol. 33, 224-227. Jamieson, G. A., Mai land, N. J., Wilcock, G. K., Ya es, C. M. and I zhaki, R. F. (1992). He pes simplex i us ype 1 DNA is p esen in speci ic egions o b ain om aged people wi h and wi hou senile demen ia o he Alzheime ype. J. Pa hol. 167, 365-368. Jellinge , K. A. (2006). Alzheime 100 –highligh s in he his o y o Alzheime esea ch. J. Neu al T ansm. 113, 1603-1623. Kell, D. B. and P e o ius, E. (2015). On he ansloca ion o bac e ia and hei lipopolysaccha ides be ween blood and pe iphe al loca ions in ch onic, in lamma o y diseases: he cen al oles o LPS and LPS-induced cell dea h. In eg . Biol. 7, 1339-1377. Kok, E., Haikonen, S., Luo o, T., Huh ala, H., Goebele , S., Haapasalo, H. and Ka hunen, P. J. (2009). Apolipop o ein E-dependen accumula ion o Alzheime disease- ela ed lesions begins in middle age. Ann. Neu ol. 65, 650-657. K is en, H., San ana, S., Sas e, I., Recue o, M., Bullido, M. J. and Aldudo, J. (2015). He pes simplex i us ype 2 in ec ion induces AD-like neu odegene a ion ma ke s in human neu oblas oma cells. Neu obiol. Aging 36, 2737-2747. Kuma , D. K., Choi, S. H., Washicosky, K. J., Eime , W. A., Tucke , S., Gho ani, J., Le kowi z, A., McColl, G., Golds ein, L. E., Tanzi, R. E. e al. (2016). Amyloid-βpep ide p o ec s agains mic obial in ec ion in mouse and wo m models o Alzheime ’s disease. Sci. T ansl. Med. 25, 1-16. Le enneu , L., Pe es, K., Fleu y, H., Ga igue, I., Ba be ge -Ga eau, P., Helme , C., O gogozo, J.-M., Gau hie , S. and Da igues, J.-F. (2008). Se oposi i i y o he pes simplex i us an ibodies and isk o Alzheime ’s disease: a popula ion- based coho s udy. PLoS ONE 3, e3637. Lin, W.-R., Wozniak, M. A., Coope , R. J., Wilcock, G. K. and I zhaki, R. F. (2002). He pes i uses in b ain and Alzheime ’s disease. J. Pa hol. 197, 395-402. Li le, C. S., Joyce, T. A., Hammond, C. J., Ma a, H., Cahn, D., Appel , D. M. and Balin, B. J. (2014). De ec ion o bac e ial an igens and Alzheime ’s disease-like pa hology in he cen al ne ous sys em o BALB/c mice ollowing in anasal in ec ion wi h a labo a o y isola e o Chlamydia pneumoniae. F on . Aging Neu osci. 6,304. Lo  heim, H., Gil ho pe, J., Johansson, A., E iksson, S., Hallmans, G. and Elgh, F. (2015a). He pes simplex in ec ion and he isk o Alzheime ’s disease: a nes ed case-con ol s udy. Alzheime s Demen . 11, 587-592. Lo  heim, H., Gil ho pe, J., Adol sson, R., Nilsson, L. G. and Elgh, F. (2015b). Reac i a ed he pes simplex in ec ion inc eases he isk o Alzheime ’s disease. Alzheime s Demen . 11, 593-599. Mahony, J. B., Luins a, K. E., Sello s, J. W. and Che nesky, M. A. (1993). Compa ison o plasmid- and ch omosome-based polyme ase chain eac ion assays o de ec ing Chlamydia achoma is nucleic acids. J. Clin. Mic obiol. 31, 1753-1758. Miklossy, J. (2011). Eme ging oles o pa hogens in Alzheime disease. Expe Re . Mol. Med. 13, e30. Miklossy, J. (2015). His o ic e idence o suppo a causal ela ionship be ween spi oche al in ec ions and Alzheime ’s disease. F on . Aging Neu osci. 7, 46. Miklossy, J., Kis, A., Radeno ic, A., Mille , L., Fo o, L., Ma ins, R., Reiss, K., Da binian, N., Da eka , P., Mihaly, L. e al. (2006). Be a-amyloid deposi ion and Alzheime ’s ype changes induced by Bo elia spi oche es. Neu obiol. Aging 27, 228-236. Pisa, D., Alonso, R., Rabano, A., Rodal, I. and Ca asco, L. (2015). Di e en b ain egions a e in ec ed wi h ungi in Alzheime ’s disease. Sci. Rep. 5, 15015. Po gie e , M., Bes e , J., Kell, D. B. and P e o ius, E. (2015). The do man blood mic obiome in ch onic, in lamma o y diseases. FEMS Mic obiol. Re . 39, 567-591. 1354 RESEARCH ARTICLE Disease Models & Mechanisms (2016) 9, 1349-1355 doi:10.1242/dmm.026674 Disease Models & Mechanisms San ana, S., Recue o, M., Bullido, M. J., Valdi ieso, F. and Aldudo, J. (2012). He pes simplex i us ype I induces he accumula ion o in acellula β-amyloid in au ophagic compa men s and he inhibi ion o he non-amyloidogenic pa hway in human neu oblas oma cells. Neu obiol. Aging 33, 430.e19-33. Shipley, S. J., Pa kin, E. T., I zhaki, R. F. and Dobson, C. B. (2005). He pes simplex i us in e e es wi h amyloid p ecu so p o ein p ocessing. BMC Mic obiol. 5,48. Smi h, J. S. and Robinson, N. J. (2002). Age-speci ic p e alence o in ec ion wi h he pes simplex i us ypes 2 and 1: a global e iew. J. In ec . Dis. 186, S3-S28. Soscia, S. J., Ki by, J. E., Washicosky, K. J., Tucke , S. M., Ingelsson, M., Hyman, B., Bu on, M. A., Golds ein, L. E., Duong, S., Tanzi, R. E. e al. (2010). The Alzheime ’s disease-associa ed amyloid be a-p o ein is an an imic obial pep ide. PLoS ONE 5, e9505. S eel, A. J. and Eslick, G. D. (2015). He pes i uses inc ease he isk o Alzheime ’s disease: a me a-analysis. J. Alzheime s Dis. 47, 351-364. Un e gasse , A., Nij een, H., Rao, X., Bisseling, T., Geu s, R. and Leunissen, J. A. (2007). P ime 3Plus, an enhanced web in e ace o P ime 3. Nucleic Acids Res. 35, W71-W74. an Duijn, C. M., de Knij , P., C u s, M., Wehne , A., Ha ekes, L. M., Ho man, A. and Van B oeckho en, C. (1994). Apolipop o ein E4 allele in a popula ion– based s udy o ea ly–onse Alzheime ’s disease. Na . Gene . 7, 74-78. Wang, J.-H., Gouda-Vossos, A., Dzamko, N., Halliday, G. and Huang, Y. (2013). DNA ex ac ion om esh- ozen and o malin- ixed, pa a in-embedded human b ain issue. Neu osci. Bull. 29, 649-654. Whi e, M. R., Kandel, R., T ipa hi, S., Condon, D., Qi, L., Taubenbe ge , J. and Ha sho n, K. L. (2014). Alzheime ’s associa ed β-amyloid p o ein inhibi s in luenza A i us and modula es i al in e ac ions wi h phagocy es. PLoS ONE 9, e101364. Whi ley, R. J. and Roizman, B. (2001). He pes simplex i us in ec ions. Lance 357, 1513-1518. Wozniak, M. A., Shipley, S. J., Comb inck, M., Wilcock, G. K. and I zhaki, R. F. (2005). P oduc i e he pes simplex i us in b ain o elde ly no mal subjec s and Alzheime ’s disease pa ien s. J. Med. Vi ol. 75, 300-306. Wozniak, M. A., Mee, A. P. and I zhaki, R. F. (2009). He pes simplex i us ype 1 DNA is loca ed wi hin Alzheime ’s disease amyloid plaques. J. Pa hol. 217, 131-138. 1355 RESEARCH ARTICLE Disease Models & Mechanisms (2016) 9, 1349-1355 doi:10.1242/dmm.026674 Disease Models & Mechanisms