Immunological Cross-Reactivity of an Ancestral and the Most Recent Pandemic Norovirus GII.4 Variant

i uses A icle Immunological C oss-Reac i i y o an Ances al and he Mos Recen Pandemic No o i us GII.4 Va ian Ki si Tamminen, Ma ia Malm, Timo Vesika i and Vesna Blaze ic * Vaccine Resea ch Cen e , Facul y o Medicine and Heal h Technology, Tampe e Uni e si y, Bioka u 10, FI-33520 Tampe e, Finland; [email p o ec ed] (K.T.); [email p o ec ed] (M.M.); [email p o ec ed] (T.V.) *Co espondence: [email p o ec ed], Tel.: +358-50-4211-054 Recei ed: 19 Decembe 2018; Accep ed: 18 Janua y 2019; Published: 22 Janua y 2019   Abs ac : No o i us (NoV) geno ype GII.4 is esponsible o he majo i y o NoV in ec ions causing pandemics e e y ew yea s. A NoV i us-like pa icle (VLP)-based accine should op imally co e he high an igenic a ia ion wi hin he GII.4 geno ype. We compa ed he immune esponses gene a ed by VLPs o he ances al GII.4 1999 s ain (GII.4 1995/96 US a ian ) and he mos ecen GII.4 Sydney 2012 pandemic s ains in mice. No signi ican di e ences we e obse ed in he ype-speci ic esponses bu GII.4 1999 VLPs we e mo e po en in inducing high-a idi y an ibodies wi h be e c oss- eac i i y. GII.4 1999 immune se a blocked binding o GII.4 2006 and GII.4 2012 VLPs o he pu a i e ecep o s in a su oga e neu aliza ion assay, whe eas GII.4 2012 immune se a only had low blocking ac i i y agains GII.4 2006 VLPs. Amino acid subs i u ion in he NERK mo i (amino acids 310, 316, 484, and 493, espec i ely), al e ing he access o conse ed blocking epi ope F, mode a ely imp o ed he c oss-blocking esponses agains mu a ed GII.4 2012 VLPs (D310N). NoV GII.4 1999 VLPs, up aken and p ocessed by an igen-p esen ing cells, induced s onge in e e on gamma (IFN- γ ) p oduc ion om mice splenocy es han GII.4 2012 VLPs. These esul s suppo he use o GII.4 1999 VLPs as a majo componen o a NoV accine. Keywo ds: no o i us; ances o ; a ian ; GII.4; immune esponses; c oss- eac i i y; blocking an ibodies; NERK mo i 1. In oduc ion No o i us (NoV) GII.4 is he p edomina ing NoV geno ype, causing up o 85% o acu e gas oen e i is ou b eaks o NoV and spo adic in ec ions wo ldwide [ 1 ]. I is associa ed wi h mo e se e e clinical mani es a ions han o he NoV geno ypes [ 2 , 3 ]. NoV genog oup II (GII) and GI oge he comp ise o o e 28 gene ically di e gen NoV geno ypes in ec ing humans [ 4 ]. The p edominance o GII.4 o o e wo decades is associa ed wi h se e al ac o s including as eplica ion and e ec i e pe son- o-pe son ansmission a es [ 5 , 6 ]. In addi ion, GII.4 a ian s ecognize a wide ange o mucosal polysaccha ides [ 7 ] and his o-blood g oup an igens (HBGAs), which a e hough o acili a e NoV en y and/o in ec ion [8,9]. The NoV pa icle consis s o 90 dime s o capsid VP1 p o ein o ganized in T = 3 icosahed al symme y [ 10 ]. VP1 is di ided in o wo main domains: he shell (S) and he p o uding (P) domains, he la e o which is u he subdi ided in o P1 and P2 domains [ 10 ]. The ou e mos P2 domain con ains he conse ed HBGA binding si es bu he egions su ounding hese si es a e e ol ing due o cons an immune p essu e [ 11 ]. NoV accine de elopmen is la gely based on NoV i us-like pa icles (VLPs) [ 12 – 14 ], an igenically iden ical o he i us pa icle, despi e ecen p og ess in cul i a ing NoV in i o [15]. Vi uses 2019,11, 91; doi:10.3390/ 11020091 www.mdpi.com/jou nal/ i uses Vi uses 2019,11, 91 2 o 15 GII.4 NoV unde goes epochal e olu ion simila o in luenza i us; pe iods o s asis lead in o apid an igenic d i in common s uc u al epi opes ha a e unde immune p essu e [ 16 , 17 ]. NoV-speci ic se um an ibodies ha block binding o NoV VLPs o HBGAs in a su oga e neu aliza ion assay a e he bes co ela e o p o ec ion iden i ied so a [ 18 – 20 ]. An igenic d i d i en by hese an ibodies can lead he e ol ing s ains gaining new HBGA binding abili ies and/o escaping om p e iously gained immuni y [ 21 – 23 ]. The e a e a leas six (A–F) e ol ing “blocking epi opes” desc ibed [ 21 , 23 – 25 ] and he eme gence o a new GII.4 pandemic s ain is ypically associa ed wi h mu a ions in hese epi opes [ 23 , 24 , 26 ]. Since he 1990s se en pandemics ha e been caused by GII.4 a ian s: G imsby (1995/96 US), Fa ming on Hills (2002), Hun e (2004), Ye seke (2006a), Den Haag (2006b), New O leans (NO, 2009), and Sydney (2012) [1]. Vi ions a e dynamic s uc u es eac ing o en i onmen wi h con o ma ional changes, which enable biologically ele an unc ions such as ecep o /ligand binding [ 27 ]. Mu a ions in he i ion co e, o neu alizing an ibody binding o ce ain epi opes, can s e ically block ecep o binding si e o cause con o ma ional change in dis an epi opes impai ing ligand in e ac ions [ 27 – 29 ]. Some o he highly a iable blocking epi opes o NoV GII.4 a e exposed on he su ace o P2-domain (e.g., epi opes A and D) while o he s, like epi ope F, a e bu ied and b oadly conse ed [ 25 ]. Epi ope F is conside ed a uni e sal GII.4 blocking epi ope, as monoclonal an ibody speci ic o epi ope F has been shown o c oss-block a panel o ime-o de ed GII.4 VLPs [ 30 ]. Amino acids 310, 316, 484, and 493 comp ise “ he NERK mo i ”, which has been sugges ed o limi an ibody access o epi ope F [31]. In his s udy we in es iga ed ype-speci ic and c oss- eac i e humo al and cellula immune esponses induced in mice wi h he i s (GII.4 1999, a 1995/96 US a ian ) and he la es (GII.4 Sydney, 2012) pandemic GII.4 a ian VLPs. In addi ion, we s udied he e ec o an amino acid mu a ion in he NERK mo i on c oss-blocking an ibody esponses. The esul s p esen ed he e add o he cu en knowledge and unde s anding o c oss- eac i e immune esponses induced in i o by di e en a ian GII.4 VLPs. 2. Ma e ials and Me hods 2.1. NoV VLPs Th ee di e en NoV GII.4 a ian VLPs we e used as immunogens and/o in i o an igens in his s udy. GII.4 1999 (o iginal pa ien sequence isola ed in 1999; i has one amino acid di e ence (aa 333) o he VP1 sequence o he e e ence s ain GII.4 1995/96 US, Genbank accession numbe AF080551) and GII.4 2012 (accession numbe AFV08795.1) VLPs we e p oduced in baculo i us-insec cell sys em and pu i ied as p e iously desc ibed by ou labo a o y [ 32 – 34 ]. GII.4 2006 a ian VLPs (accession numbe BAG70446) we e p oduced by Icon Gene ics Gmbh (Halle, Ge many) [ 35 ] and we e u ilized in analy ical me hods only. The epi ope-enginee ed GII.4 2012 pFas Bac1- ec o wi h amino acid D310 subs i u ed o N310 was o de ed om GeneA (The mo Fishe Scien i ic, Wal ham, MA, USA) and exp essed using he Bac- o-Bac Baculo i us Exp ession Sys em (In i ogen, Ca lsbad, CA, USA) as desc ibed be o e [ 32 ]. Mu an VLPs ( e e ed as GII.4 2012 D310N) we e p oduced and pu i ied wi h he same me hodology as he wild- ype VLPs. The iden i y, pu i y, and mo phology [ 36 , 37 ] o he VLPs we e con i med as desc ibed elsewhe e [32,33,38]. 2.2. Mouse Immuniza ions and Tissue Collec ions Se en-week-old emale BALB/c OlaHsd mice ob ained om En igo RMS BV (Ho s , he Ne he lands) we e immunized in amuscula ly (IM) wice (weeks 0 and 3) wi h 10 µ g o GII.4 1999 (5 mice) o GII.4 2012 VLPs (4 mice). Mice ecei ing ca ie bu e only (phospha e bu e ed saline, PBS) we e used as nega i e con ols (5 mice). Whole blood and lymphoid issue we e collec ed a he ime o eu haniza ion on s udy week 5. Fo bone ma ow-de i ed dend i ic cell (BMDC) gene a ion, emu s and ibia we e collec ed om nai e con ol mice and he ex e io issues we e s e ilized wi h 70% e hanol and kep on ice un il bone ma ow ex ac ion (desc ibed in Sec ion 2.6). Se um was sepa a ed by Vi uses 2019,11, 91 3 o 15 cen i uga ion and splenocy e suspensions o each mice we e p epa ed acco ding o ea lie published me hods [ 38 ]. All p ocedu es we e au ho ized and pe o med in conco dance wi h he guidelines by he Finnish Na ional Animal Expe imen Boa d (pe mission numbe ESAVI/10800/04.10.07/2016). 2.3. IgG Ti e and A idi y Assay An igen-speci ic and c oss- eac i e immunoglobulin G (IgG) esponses in mice se a we e measu ed by enzyme-linked immunoso ben assay (ELISA) as desc ibed in de ail elsewhe e [ 32 , 38 ]. Indi idual se a we e added by dec easing wo- old dilu ions (IgG i e de e mina ion) o 1:100 dilu ion (a idi y assay) on 96-well hal -a ea polys y ene pla es (Co ning Inc., Co ning, NY, USA) coa ed o e nigh a 4 ◦ C wi h NoV VLPs (1 µ g/mL) and blocked o one hou a oom empe a u e (RT) wi h 2% skimmed milk in PBS/0.05%Tween. A e se um incuba ion, he bound an ibodies we e de ec ed by goa an i-mouse IgG-HRP (dilu ion 1:6000, Sigma-Ald ich, Sain Louis, MO, USA) eac ing wi h o-phenylenediamine dihyd ochlo ide (OPD)-subs a e (30 min RT). All incuba ions we e pe o med a 37 ◦ C o one hou unless o he wise s a ed. A e s opping he subs a e eac ion wi h 2M H 2 SO 4 he op ical densi y (OD) was measu ed a 490 nm (Vic o 2 1420; Pe kinElme , Wal ham, MA, USA). Each sample/dilu ion was assayed in duplica e wells and sample olumes we e 50 µ L/well. The backg ound signal om wells lacking se um (blank wells) was sub ac ed om all o he OD eadings a a pla e. The cu -o alue was calcula ed as mean OD + 3 × SD o nega i e con ol mice se a a a gi en dilu ion. A sample esul ing in an OD alue abo e he se cu -o OD and a leas 0.100 OD was conside ed posi i e. End-poin an ibody i e s we e de ined as he highes dilu ion o se um gi ing an OD abo e he se cu -o alue. Geome ic mean i e s (GMTs) wi h 95% con idence in e als (CIs) o each immuniza ion g oup we e coun ed om indi idual mice end-poin i e s. Se um IgG a idi y was measu ed by ELISA as desc ibed abo e, bu a e se um incuba ion he pla es we e incuba ed wice ( o 5 min o each ea men ) wi h 8 M u ea (Sigma-Ald ich) o elu e low-a idi y an ibodies [39]. The a idi y index was calcula ed as (OD wi h u ea/OD wi hou u ea) ×100%. 2.4. Ca bohyd a e Binding Assays HBGA binding o mu a ed GII.4 2012 D310N VLPs in compa ison o wild- ype GII.4 2012 VLPs was analyzed by using h ee di e en sou ces o HBGAs: human ype A sali a [ 7 , 40 ], pig gas ic mucin (PGM, ype III, Sigma-Ald ich) [41,42], and syn he ic bio inyla ed HBGAs (Lewisa, Lewisb, H- ype-1, H- ype-3, A- ime , and B- ime , Glyco ech, Gai he sbu g, MD, USA) [ 7 , 18 ]. B ie ly, sali a (a dilu ion 1:3000, o/n 37 ◦ C) o PGM (2.5 µ g/mL, o/n RT) we e used o coa hal -a ea 96-well pla es (Co ning) and syn he ic bio inyla ed ca bohyd a es (2.5 µ g/mL, 1 h RT) we e used on p ecoa ed Neu A idin pla es (Pie ce, Rock o d, MI, USA). VLPs we e added on he pla es wi h inal concen a ion o 0.1 µ g/mL o 1.5 h (sali a and PGM pla es) o 0.4 µ g/mL o 2 h (syn he ic HBGA pla es). The bound VLPs we e de ec ed wi h human GII.4-posi i e se um ollowing ho se adish pe oxidase (HRP)-conjuga ed an i-human IgG seconda y an ibody (No ex, In i ogen) incuba ion (1 h). A e coa ing, sali a pla es we e incuba ed a 37 ◦ C, PGM pla es a RT, and syn he ic HBGA pla es a 4 ◦ C acco ding o he empe a u es used in he o iginal publica ions [ 18 , 40 , 41 ]. The pla es we e de eloped and s opped as desc ibed in Sec ion 2.3 o ELISA p ocedu es. Posi i e eac i i y is de ined as a mean OD >0.2 a e backg ound sub ac ion. Each sample was es ed in 2–4 eplica e wells and esul s a e ep esen ed as mean OD om pa allel wells. 2.5. Blocking Assays Blocking assay was used o measu e an ibodies ha block binding o NoV VLPs o HBGAs p esen in human sali a ( ype A) o PGM, o o syn he ic bio inyla ed H- ype-1 ca bohyd a e as p e iously desc ibed [ 7 , 42 ]. NoV VLPs wi h inal concen a ion o 0.1 µ g/mL (sali a and PGM assays) o 0.4 µ g/mL (syn he ic HBGA assays) we e p eincuba ed o 1 h a 37 ◦ C in low binding ubes wi h dec easing concen a ion o mice se um (s a ing a 1:100 se um dilu ion o homologous and 1:20 o c oss-blocking assays). The p e-incuba ed VLP-se um mix u es we e hen added on sali a, PGM, Vi uses 2019,11, 91 4 o 15 o syn he ic H- ype-1 coa ed pla es, and he bound VLPs we e de ec ed wi h human GII.4-posi i e se um ( o GII.4 1999 and GII.4 2012 VLP de ec ion) o abbi NoV-hype immune se um ( o GII.4 2006 VLP de ec ion) ollowing he co esponding HRP-conjuga ed seconda y an ibody incuba ion. The incuba ion imes and empe a u es we e he same as desc ibed in Sec ion 2.4 o VLP-binding assays. The pla es we e de eloped, s opped, and measu ed as desc ibed in Sec ion 2.3. Wells incuba ed wi h VLPs lacking mouse se a we e used o de e mine he maximum binding OD. The blocking index (%) was calcula ed as 100% − [(OD wells wi h VLP − se um mix/OD maximum binding OD) × 100%]. The blocking i e 50 (BT50) alue ep esen s he highes se um dilu ion blocking 50% o he maximum VLP binding. The esul s a e exp essed as he mean blocking indexes o indi idual mice o he mean o eplicas o epea ed expe imen s i pooled g oup se a we e used. 2.6. BMDC Gene a ion and Pulsing The me hod o gene a ing mouse BMDCs was adap ed om a published p ocedu e [ 43 ] wi h some modi ica ions. A e emo ing so issue, he emu s and he ibiae we e cu om each end wi h scalpel and lushed wi h ice cold PBS. The ex ac ed bone ma ows we e passed h ough a 70- µ m cell s aine (Bec on-Dickinson, BD, F anklin Lakes, NJ, USA) and collec ed in comple e medium (CM, RPMI-1640 supplemen ed wi h 100 U/mL penicillin, 100 µ g/mL s ep omycin, 50 µ m 2-me cap oe hanol, 2 mm l-glu amine, and 10% e al bo ine se um (FBS), all pu chased om Sigma-Ald ich). The cell suspensions we e cen i uged 300 × g o 10 min and suspended in CM con aining 20 ng/mL ecombinan mouse g anulocy e-mac ophage colony-s imula ing ac o (GM-CSF, Abcam, Camb idge, UK). BM-cells we e seeded a 1 × 10 6 cells/mL (10 mL pe pla e) in non- ea ed 90 ×14.2-mm s e ile pe i dishes (VWR, Radno , PA, US) and cul u ed a 37 ◦ C, 5% CO 2 o 8 days. F esh CM wi h GM-CSF (5 mL/pla e) was added on he dishes on days 4 and 7 and he cells we e ha es ed on day 8. The gene a ed cells we e su ace s ained wi h phycoe y h in (PE)-conjuga ed an i-mouse CD11c and Ho izon Viabili y S ain 780 (bo h om BD) and acqui ed using BD FACSCan o II low cy ome e as desc ibed ea lie [ 44 ] which con i med he cells o be >90% CD11c+ cells. The BMDCs we e ozen acco ding o published p ocedu e [ 45 ] in ice-cold CM con aining 10% DMSO (Sigma-Ald ich). The BMDCs we e hawed, washed wice (300 × g, 10 min) and seeded 2 × 10 6 cells/mL in CM in non- ea ed cell-cul u e 24-well pla es (Co ning Cos a ) o an igen pulsing. GII.4 1999, GII.4 2012 and GII.4 2006 VLPs we e added o he cells a 100 µ g/mL. Cells lacking pulsing an igen (un-pulsed BMDCs) we e used as a nega i e con ol. The cells we e incuba ed (37 ◦ C, 5% CO 2 ) o 20–22 hou s, washed wice, and we e used as an igen-p esen ing cells (APCs) in enzyme-linked immunospo (ELISPOT) assay. 2.7. ELISPOT-In e e on Gamma (IFN-γ) Homologous and c oss- eac i e T-cell esponses we e analyzed by quan i ica ion o IFN- γ p oduc ion om g oup-wise pooled mice splenocy es in esponse o VLP pulsed BMDCs by an ELISPOT-IFN- γ assay [ 36 , 38 ]. B ie ly, 96-well Mul iSc een HTS-IP il e pla es (Millipo e, Bille ica, MA, USA) we e coa ed o e nigh a 4 ◦ C wi h an i-mouse IFN- γ (Mab ech Ab, Nacka S and, Sweden). A e washing and blocking he pla es wi h 10% FBS in CM (2–3 h RT), GII.4 1999, GII.4 2012, GII.4 2006 VLP pulsed o un-pulsed BMDCs (5000, 20,000, and 40,000 BMDCs/well) we e added on pla es. Concana alin A (ConA; Sigma-Ald ich) a 10 µ g/ml was used as a posi i e con ol o s imula e IFN- γ p oduc ion om splenocy es. GII.4 1999 o GII.4 2012 VLP-immunized mice splenocy es we e hawed, washed, and added o he pla es (0.2 × 10 6 cells/well) on op o BMDCs and he pla es we e incuba ed o 20 h a 37 ◦ C and 5% CO 2 . The ea e , he cells we e disca ded and he pla es we e de eloped wi h bio inyla ed an i-mouse IFN- γ monoclonal an ibody (0.5 µ g/mL in PBS/0.5% FBS, 2 h a RT) and alkaline-phospha ase (ALP) conjuga ed s ep a idin (1:1000, 1 h a RT) eac ing wi h BCIP/NBT subs a e (all om Mab ech). A e 12 minu es he colo eac ion was s opped wi h ap wa e . The spo s Vi uses 2019,11, 91 5 o 15 we e coun ed by an ImmunoSpo ® au oma ic CTL analyze (CTL-Eu ope GmbH, Bonn, Ge many) and he esul s a e exp essed as mean spo o ming cells (SFCs) pe 10 6 li e splenocy es o eplica e wells. 2.8. S a is ics The K uskal–Wallis’ es was used o assess he s a is ical di e ences in an ibody i e s and a idi y indices be ween indi idual immuniza ion g oups. A s a is ically signi ican di e ence was de ined as ap- alue o <0.05. Da a we e analyzed wi h IBM SPSS S a is ics e sion 25.0 (SPSS Inc., Chicago, IL, USA). 3. Resul s 3.1. NoV GII.4 Type-Speci ic and C oss-Reac i e IgG An ibody Ti e s and A idi y Immunized mice se a we e es ed using ELISA o quan i y ype-speci ic and c oss- eac i e IgG an ibodies agains GII.4 1999, GII.4 2006, and GII.4 2012 VLPs (Figu e 1a). GII.4 1999 and GII.4 2012 immuniza ions esul ed in equally high magni udes (p= 0.264) o ype-speci ic IgG esponse, wi h GMTs o 102,400 and 86,100 (95% CI = 53,200–139,300), espec i ely. IgG esponses agains homologous VLPs we e signi ican ly highe (p< 0.05) han c oss- eac i e esponses induced by he e ologous an igen. GII.4 1999 VLP immuniza ion induced signi ican ly highe (p= 0.018) c oss- eac i e IgG esponse agains GII.4 2006 VLPs (GMT 25,600, 95% CI = 9600–69,900) han GII.4 2012 VLP immuniza ion (GMT 3200, 95% CI = 1500–7000). When GII.4 1999 and GII.4 2012 c oss- eac i e esponses we e compa ed agains each o he , GII.4 1999 immuniza ion esul ed in 2- old highe GII.4 2012-speci ic i e (GMT 16,900, 95% CI = 5800–49,200) han GII.4 2012 immuniza ion agains GII.4 99 (GMT 8060, 95% CI = 3900–16,900) bu he di e ence was no s a is ically signi ican (p= 0.15). Con ol mice did no de elop speci ic IgG esponse o any o he VLPs es ed. Figu e 1. Ti e s and a idi y o no o i us (NoV) ype-speci ic and c oss- eac i e se um immunoglobulin G (IgG) an ibodies. Mice we e immunized wi h GII.4 1999 (5 mice) and GII.4 2012 (4 mice) i us-like pa icles (VLPs) and he immune se a was used in enzyme-linked immunoso ben assay (ELISA) o de e mine he magni ude o IgG an ibodies agains homologous and he e ologous NoV VLPs ( a ). Se um o mice ecei ing phospha e bu e ed saline (PBS) (5 mice) was used as a nega i e con ol (C l). Shown a e he geome ic mean i e s (GMTs) wi h 95% con idence in e als (e o ba s) coun ed om indi idual mice end-poin i e s in each immuniza ion g oup. The dashed line illus a es he cu -o i e o samples conside ed posi i e. The a idi y o IgG an ibodies was measu ed om indi idual mice se a agains homologous and he e ologous NoV VLPs ( b ) as desc ibed in he Ma e ial and Me hods. Ho izon al lines in he box plo s ep esen he medians, c oss-symbols ( × ) ep esen he means, and he boxes illus a e he in e qua ile ange ha con ains 50% o alues wi h whiske s ex ending o he highes and lowes alues. The an igen-speci ic an ibody i e s and he a idi y indexes be ween immuniza ion g oups we e compa ed by he K uskal–Wallis es and signi ican di e ences (p alue <0.05) a e iden i ied wi h an as e isk (*). Vi uses 2019,11, 91 6 o 15 Compa ison o he a idi y o ype-speci ic and c oss- eac i e an ibodies a e shown in Figu e 1b. A ema kable di e ence was obse ed in he ype-speci ic a idi ies elici ed by GII.4 1999 and GII.4 2012 VLPs (Figu e 1b). GII.4 1999 VLPs induced ype-speci ic IgG an ibodies wi h high a idi y (mean a idi y index 85.7 ± 9%), whe eas GII.4 2012 ype-speci ic an ibody a idi y was poo (19.4 ± 10.7%). As expec ed, he a idi y o GII.4 1999 immune se um agains he e ologous VLPs was conside ably lowe han ype-speci ic a idi y, bu s ill a in e media e le el (mean a idi y indexes anging om 30% o 40%). GII.4 1999 VLP immuniza ion elici ed c oss- eac i e an ibodies agains GII.4 2006 VLPs wi h signi ican ly (p= 0.027) highe a idi y han GII.4 2012 VLP immuniza ion. 3.2. Blocking An ibody Responses The abili y o mouse immune se a o block homologous and he e ologous VLPs binding o HBGAs p esen in human sali a ( ype A) was es ed in a blocking assay [ 7 , 40 ]. Bo h GII.4 1999 VLP and GII.4 2012 VLP immuniza ions induced s ong homologous blocking ac i i y as mean se um i e s blocking a leas 50% o VLP binding (BT50) we e 1:800 o GII.4 1999 and 1:400 o GII.4 2012 (p= 0.079, Figu e 2a). Howe e , when GII.4 1999 and GII.2 2012 immune se a we e used o block he e ologous VLPs binding, signi ican di e ences we e obse ed in c oss-blocking esponses (Figu e 2b,c). GII.4 1999 VLP immuniza ion (Figu e 2b) elici ed s ong blocking ac i i y agains GII.4 2006 VLPs (BT50 = 1:640) while blocking agains GII.4 2012 VLP binding was conside ably lowe (BT50 = 1:40). In con as , GII.4 2012 immune se a (Figu e 2c) blocked only GII.4 2006 VLP binding (BT50 = 1:40) a a signi ican ly lowe magni ude han GII.4 1999 immune se a (p= 0.036) and comple ely ailed o block GII.4 1999 binding. Con ol mice se a blocked all VLP binding <15% a he lowes se um dilu ion (1:20). Figu e 2. No o i us (NoV) ype-speci ic and c oss- eac i e blocking an ibodies. GII.4 1999 (5 mice) o GII.4 2012 (4 mice) i us-like pa icle (VLP) immunized mice se a we e indi idually dilu ed 2- old s a ing om 1:100 dilu ion and assayed o he blocking o homologous NoV VLPs binding o human ype A sali a ( a ). The c oss-blocking ac i i y o GII.4 1999 ( b ) and GII.4 2012 ( c ) VLP immunized mice se a as well as con ol mice (C l, 5 mice) was assayed agains he e ologous VLPs as 2- old dilu ion se ies s a ing om 1:20 dilu ion. The blocking index (pe cen ) was calcula ed as ollows: 100% × ((OD wells wi h se um/OD wells wi hou se um, maximum binding) × 100%). The symbols ep esen he immuniza ion g oup mean blocking indexes and he e o ba s ep esen he s anda d e o be ween indi idual mice. The ho izon al dashed line ep esen s he blocking i e 50% (BT50). 3.3. Mo phology, An igenici y, and HBGA-Binding P o ile o Gene ically Enginee ed GII.4 2012 D310N VLPs The NERK mo i (amino acids 310, 316, 484, and 493, espec i ely) is sugges ed o con ol an ibody access o epi ope F, a pu a i e uni e sal blocking epi ope in he GII.4 lineage [ 31 ]. We gene ically enginee ed GII.4 2012 VLP o e e amino acid 310 om D o N, as p esen in GII.4 1999 VLP. The mo phology o GII.4 2012 D310N VLPs was s udied by elec on mic oscopy, which con i med ha he mu a ion did no a ec mu a ed VLP in eg i y o mo phology (Figu e 3a) in compa ison o wild ype GII.4 2012 VLPs (Figu e 3b). The an igenici y o GII.4 2012 D310N VLPs was u he in es iga ed using ELISA (Figu e 3c). GII.4 1999 and GII.4 2012 VLP-immunized mice se a ecognized wild- ype Vi uses 2019,11, 91 7 o 15 (GII.4 2012) and mu a ed VLPs (GII.4 2012 D310N) wi h simila in ensi y (Figu e 3c). HBGA binding assay agains human sali a A, PGM, and syn he ic ca bohyd a es con i med ha mu a ion D310N did no a ec he ligand-binding abili ies o he mu a ed VLPs (Figu e 3d). Figu e 3. Mo phology, an igenici y, and his o-blood g oup an igen (HBGA) binding o wild- ype and mu a ed GII.4 2012 i us-like pa icles (VLPs). Mo phology o GII.4 2012 D310N ( a ) and GII.4 2012 ( b ) VLPs we e examined by FEI Tecnai F12 elec on mic oscope (Philips 487 Elec on Op ics, Holland) a e nega i e s aining wi h 3% u anyl ace a e pH 4.6. Mu a ed GII.4 2012 D310N VLPs and wild- ype GII.4 2012 VLPs we e used as an igens wi h enzyme-linked immunoso ben assay (ELISA), eac ing wi h GII.4 2012, GII.4 1999, and con ol mice (c l) immune se a ( c ). VLPs we e assayed o binding o sali a ype A, pig gas ic mucin (PGM), and syn he ic his o-blood g oup an igens as desc ibed in he Ma e ials and Me hods ( d ). The symbols ( c ) and ba s ( d ) illus a e he mean op ical densi y (OD) a 490 nm o eplica e wells (2–4) wi h s anda d e o s. Le a , Lewis a ; Le b , Lewis b ; H- -1, H- ype-1; H- -3, H- ype 3. Vi uses 2019,11, 91 8 o 15 3.4. Blocking An ibody Responses agains Gene ically Enginee ed GII.4 2012 D310N VLPs To s udy he e ec o amino acid 310 mu a ion on c oss-blocking ac i i y we used GII.4 1999 pooled immune se a o block wild- ype GII.4 2012 and mu a ed GII.4 2012 D310N VLP binding o ype A sali a (Figu e 4a), PGM (Figu e 4b), and syn he ic H- ype-1 (Figu e 4c). In sali a A and H- ype-1 blocking assays BT50 inc eased 2- old ( om 1:40 o 1:80, Figu e 4a and om 1:80 o 1:160, Figu e 4c), and in PGM blocking assay 4- old ( om 1:40 o 1:160, Figu e 4b) in a o o he mu a ed VLPs. We also in es iga ed he se a blocking ac i i y o GII.4 2012-immunized mice agains GII.4 2012 D310N mu an in syn he ic H- ype-1 blocking assay (Figu e 4d) and no di e ences we e obse ed in blocking o wild- ype and mu a ed GII.4 2012 VLPs. Figu e 4. The e ec o amino acid (aa) 310 mu a ion wi hin he NERK mo i on c oss-blocking ac i i y. Residue 310 o GII.4 2012 i us-like pa icles (VLPs) was subs i u ed om D (aspa ic acid) o N (aspa agine) o gene a e mu a ed GII.4 2012 D310N VLPs. Pooled GII.4 1999 immune se um was assayed agains wild- ype and mu a ed GII.4 2012 VLPs in blocking assays u ilizing human sali a ype A ( a ), pig gas ic mucin ( b ), o syn he ic H- ype-1 ( c ) as he sou ce o his o-blood g oup an igens (HBGAs). Pooled GII.4 2012 immune se um was used o block he binding o GII.4 2012 and GII.4 2012 D310N VLPs o syn he ic H- ype-1 HBGAs ( d ). Con ol (C l) mice se a illus a e he non-speci ic blocking ac i i y. The blocking index (pe cen ) was calcula ed as ollows: 100% × ((OD wells wi h se um/OD wells wi hou se um, maximum binding) × 100%). The symbols ep esen he mean blocking indexes wi h s anda d e o s be ween wo epea ed assays (( a ) and ( c )) o duplica e wells (( b ) and (d)) and he ho izon al dashed line ep esen s he blocking i e 50% (BT50). 3.5. T-Cell Responses ELISPOT IFN- γ was used o in es iga e whe he he e a e di e ences in he de elopmen o homo ypic o he e o ypic T-cell immuni y a e GII.4 1999 and GII.4 2012 VLP immuniza ion. IFN- γ -p oducing cells we e quan i ied om mice splenocy es in esponse o au ologous BMDCs pulsed wi h di e en NoV GII.4 VLPs. GII.4 1999 o GII.4 2012 VLP-immunized mice splenocy es esponded wi h simila in ensi y o GII.4 1999 (Figu e 5a), GII.4 2012 (Figu e 5b) and GII.4 2006 VLP (Figu e 5c) pulsed BMDCs used as APCs in he ELISPOT. The highes IFN- γ p oduc ion was induced wi h GII.4 1999 VLP pulsed APCs ollowed by GII.4 2012 and GII.4 2006 VLP pulsed cells. The magni ude o IFN- γ eleasing cells inc eased wi h highe numbe o pulsed BMDCs. No signi ican Vi uses 2019,11, 91 9 o 15 IFN-γp oduc ion was de ec ed when un-pulsed BMDCs we e used as a nega i e con ol (Figu e 5d). All samples we e es ed in duplica e cells. Cell iabili y o he esponding cells was simila in all assays as con olled by Con A s imula ion (da a no shown). Backg ound con ol (splenocy es in CM only) esul ed in <40 spo s pe 106cells. Figu e 5. No o i us (NoV) GII.4-speci ic T-cell esponses. G oup-wise pooled splenocy es o mice immunized wi h GII.4 1999 i us-like pa icles (VLPs), GII.4 2012 VLPs, o PBS only (con ol mice) we e analyzed o in e e on gamma (IFN- γ ) p oduc ion by an ELISPOT assay. Inc easing numbe (legend) o bone ma ow-de i ed dend i ic cells (BMDCs) pulsed wi h GII.4 1999 ( a ), GII.4 2012 ( b ), o GII.4 2006 ( c ) VLPs we e used o s imula e he splenocy es o GII.4 1999 o GII.4 2012 VLP-immunized mice. Un-pulsed BMDCs ( d ) se ed as a nega i e con ol. Mean IFN- γ spo - o ming cells pe 10 6 li e splenocy es o duplica e wells wi h s anda d de ia ions (e o ba s) a e shown. 4. Discussion An impo an issue in NoV VLP accine de elopmen is he an igenic di e si y o NoV geno ypes and he e olu ion o he p edominan GII.4 s ain esul ing in a ian s able o escape he d immuni y [ 5 ]. A simila phenomenon d i es he an igenic d i o in luenza i us, and he e o e in luenza accine mus be e o mula ed on a yea ly basis o ma ch he ci cula ing s ains [ 16 , 46 ]. The NoV VLP accine migh also need o be e o mula ed e e y ew yea s, o else c oss-blocking epi opes could be induced o gene a e b oadly blocking an ibodies ha p o ec agains a a ie y o NoV a ian s [ 28 , 30 , 47 ]. In ou ea lie s udies we ha e shown ha ances al GII.4 1999 VLPs end o induce immune esponses wi h be e c oss- eac i i y and highe quali y han o he NoV geno ype VLPs [ 48 – 50 ]. To u he in es iga e GII.4 1999 VLP po en ial as a accine an igen, we compa ed he immune esponses induced by GII.4 1999 VLPs and he mos ecen pandemic a ian GII.4 2012 VLPs in mice. GII.4 2006 VLPs, ep esen ing pandemic a ian 2006a, we e selec ed as a he e ologous GII.4 an igen as he gene ic dis ances o GII.4 1999 and GII.4 2012 a ian s we e app oxima ely equal (Figu e 6a), wi h VP1 sequence amino acid iden i ies o 94.4% and 94.1%, espec i ely (Figu e 6b). A idi y o an ibodies is conside ed o be an impo an su oga e o p o ec i e e icacy o se e al accines [ 51 , 52 ] and high a idi y enhances he c oss- eac i i y o an ibodies by ole a ing mino a ia ion in he a ge epi opes [ 53 ]. In his s udy, only GII.4 1999 VLPs we e able o induce ype-speci ic an ibodies wi h high a idi y, whe eas GII.4 2012 VLPs induced only an ibodies wi h e y low a idi y. B cells ha ake up an an igen can ei he ma u e in o ex a ollicula plasmablas s sec e ing low-a ini y an ibodies o en e in o ge minal cen e (GC) whe e a ini y ma u a ion akes place wi h he help o ollicula DCs and T-cells [ 54 ]. GII.4 1999 and GII.4 2012 VLPs may be di e en ly up aken o p esen ed in GC a ec ing a ini y ma u a ion, bu u he s udies a e needed o con i m his no ion. Howe e , based on he esul s ob ained he e, an ibody a idi y canno be conside ed as a single co ela e o a s ong blocking ac i i y, as GII.4 2012 VLPs induced poo a idi y an ibodies bu s ill con e ed homologous blocking simila ly o GII.4 1999 se um. Ins ead, high an ise um a idi y may enhance he c oss-neu aliza ion abili y as b oadly neu alizing an i i al an ibodies a e usually de ec ed in ecu en /ch onic in ec ions o a e epea ed accina ions [55]. We obse ed ha GII.4 1999 immune se um had blocking ac i i y agains GII.4 2012 bu no ice e sa. The inding is in conco dance wi h ou ea lie esul s whe e GII.4 1999 immune se um was able o block he VLP binding o ano he con empo a y a ian (GII.4 NO) o HBGAs whe eas GII.4 NO