bioRxiv preprint doi: https://doi.org/10.1101/2024.11.08.622746; this version posted November 10, 2024.

The copyright holder for this preprint

(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
available under aCC-BY-NC-ND 4.0 International license.

1 Neutralizing Activity and Viral Escape of Pemivibart by

2 SARS-CoV-2 JN.1 sublineages
3

4 Tianjiao Yao1,#, Zhenghai Ma2,#, Ke Lan1,*, Yicheng Guo3,*, Lihong Liu1,*

5
1
6 State Key Laboratory of Virology, College of Life Sciences, Wuhan University, Wuhan, China
2
7 The College of Life Science and Technology, Xinjiang University, Urumqi, Xinjiang, China
3
8 Aaron Diamond AIDS Research Center, Columbia University Vagelos College of Physicians and
9 Surgeons, New York, NY 10032, USA
10
#
11 These authors contributed equally
12 *Correspondence: [email protected] (K.L.), [email protected] (Y.G.),
13 [email protected] (L.L.)
14
 bioRxiv preprint doi: https://doi.org/10.1101/2024.11.08.622746; this version posted November 10, 2024. The copyright holder for this preprint
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
available under aCC-BY-NC-ND 4.0 International license.

15 Abstract
16
17 Pemivibart (Pemgarda™/VYD222) was granted Emergency Use Authorization (EUA) by the U.S.
18 Food and Drug Administration (FDA) on March 22, 2024, for COVID-19 pre-exposure
19 prophylaxis in immunocompromised individuals. However, its efficacy and resistance against JN.1
20 sublineages have yet to be fully characterized. Here, we first assessed the neutralizing activity of
21 Pemivibart against a panel of VSV-based pseudoviruses representing contemporary JN.1
22 sublineages, including XEC, the fastest-growing SARS-CoV-2 strain globally, in both Vero-E6 and
23 Vero-E6-TMPRSS2-T2A-hACE2 (Vero-E6-TA) cells. We then engineered a replication-
24 competent vesicular stomatitis virus with JN.1 spike (rVSVΔG-JN.1) to select for escape variants
25 and performed structural analyses to comprehensively map Pemivibart’s escape mutations. Our
26 results demonstrated that Pemivibart exhibited comparable neutralization patterns in both cell lines
27 and retains broad effectiveness against SARS-CoV-2 JN.1 sublineages tested. However, its
28 potency was remarkably reduced against KP.3.1.1 and XEC, with IC50 values of approximately 4.2
29 µg/mL, about 22-fold higher than that for JN.1, as well as JN.1-derived Pemivibart-escape mutants
30 harbouring low-frequency mutations across SARS-CoV-2 strains through mutiple antibody
31 evasion mechanisms in Vero-E6-TA cells. Collectively, our findings underscored the importance
32 of monitoring the clinical efficacy of Pemivibart as JN.1 sublineages continue to evolve. The
33 escape profile of Pemivibart could provide valuable insights for forecasting and optimizing its
34 effectiveness against emerging SARS-CoV-2 variants.
35
36 Key words: SARS-CoV-2; JN.1 sublineages; Pemivibart; Neutralizing activity; Escape mutations
37
 bioRxiv preprint doi: https://doi.org/10.1101/2024.11.08.622746; this version posted November 10, 2024. The copyright holder for this preprint
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
available under aCC-BY-NC-ND 4.0 International license.

38 Introduction
39
40 Pemivibart (Pemgarda™/VYD222), a monoclonal antibody developed by Invivyd Inc., received
41 U.S. FDA emergency use authorization in March 2024 for COVID-19 pre-exposure prophylaxis
42 in moderately to severely immunocompromised individuals aged 12 and older, weighing at least
43 40 kilograms. It is the only FDA-approved antibody in clinical use with broad neutralizing activity
44 against SARS-CoV-2 variants, including Omicron JN.11,2.
45
46 Pemivibart is an optimized successor to Adintrevimab (ADG-20), targeting the class 1/4 region of
47 the receptor-binding domain (RBD) and featuring Met428Leu and Asn434Ala (LA) substitutions
48 to enhance in vivo longevity (Figure S1A and S1B). Nine additional mutations were incorporated
49 into the ADG-20 variable regions to counter specific spike escape mutations, enabling Pemivibart
50 to overcome limitations against early Omicron subvariants and regain efficacy against BQ and
51 subsequent variants (Figure S1C, S1D, and S1E). However, recent studies indicate that
52 Pemivibart is less effective against currently circulating SARS-CoV-2 JN.1 sublineages,
53 particularly KP.3.1.1, due to two spike mutations: the S31 deletion in the N-terminal domain (NTD)
54 and the Q493E mutation in the RBD3,4. Now, the JN.1 sublineage XEC, which has acquired unique
55 mutations in the NTD (T22N and F59S) compared to KP.3, is the fastest-growing variant globally
56 and gradually replacing the dominant KP.3.1.1 sublineage (Figure 1A and S2). However, it
57 remains unclear whether Pemivibart retains neutralization efficacy against XEC and how readily
58 SARS-CoV-2 JN.1 sublineages might evade it.
59
60 Results
61
62 First, we evaluated the neutralization capacity of Pemivibart against JN.1 sublineages, including
63 KP.3.1.1 and XEC, as well as pseudoviruses carrying unique point mutations from these variants
64 in both Vero-E6 and Vero-E6-TMPRSS2-T2A-hACE2 (Vero-E6-TA) cells, in comparison with
65 that of SA55 (BD55-5514), a therapeutic antibody under clinical development in China5 (Figure
66 1A). The Vero-E6-TA cell line highly expresses both TMPRSS2 and human ACE2 (hACE2) to
67 facilitate viral entry and exhibited a consistant neutralization pattern similar to Vero-E6 cells
68 (Figure 1A). Notably, SA55 retained its potency against all tested viruses. Consistent with reported
69 findings3,4, Pemivibart showed slightly reduced potency against JN.1 sublineages KP.2.3, LB.1,
70 KP.3, and XDV.1 (JN.1-F456L), but demonstrated a substantial reduction of neutralization against
71 KP.3.1.1 and XEC, with IC50 values of approximately 4.2 µg/mL, ~22-fold higher than that against
72 JN.1 in Vero-E6-TA cells. It’s worth noting that clinical studies showed a trough serum
73 concentration of Pemivibart at 175 µg/mL 90 days post-administration, which remains well above
 bioRxiv preprint doi: https://doi.org/10.1101/2024.11.08.622746; this version posted November 10, 2024. The copyright holder for this preprint
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
available under aCC-BY-NC-ND 4.0 International license.

74 the IC50 values against KP.3.1.1 and XEC1. Additionally, analysis of the neutralization profile of
75 viruses with individual mutations from JN.1 sublineages revealed that reduced sensitivity to
76 Pemivibart was primarily attributed to the S31 deletion, F59S, and Q493E, each conferring
77 approximately 3.4- to 4.5-fold increased resistance across the two cell lines tested (Figure 1A).
78
79 To investigate SARS-CoV-2 escape from Pemivibart, we employed a replication-competent
80 vesicular stomatitis virus encoding the spike protein from the SARS-CoV-2 JN.1 variant
81 (rVSVΔG-JN.1). Observing incomplete neutralization at 10 µg/ml of Pemivibart in Vero-E6-TA
82 cells (Figure S3A), we initiated viral escape selection at 20 µg/ml, followed by three rounds of
83 Pemivibart selection at a more stringent concentration of 50 µg/ml. As a result, we identified 55
84 escape variants with 43 being distinct, originating from 13 NTD single-point mutations, 22 RBD
85 single-point mutations, and 6 double mutations, selected from 480 wells of rVSVΔG-
86 JN.1/Pemivibart cultures (Figure 1B and S3B). Notably, most of these mutations occurred at rare
87 frequencies in historical and circulating SARS-CoV-2 populations, while five mutations, including
88 S31P, Y200C, R273K, P521S, and T547I, appeared in multiple variants (Table S1). Interestingly,
89 most escape mutations severely impaired JN.1 infectivity on Vero-E6 cells, yet demonstrated high
90 fitness on Vero-E6-TA cells, except for Y501D and H505R. These findings suggested that high
91 TMPRSS2 and hACE2 may facilitate viral escape by enhancing SARS-CoV-2 infection under
92 Pemivivart selection (Figure S4). We then validated these escape mutations for their ability to
93 confer resistance to Pemivibart, with SA55 and hACE2 serving as controls in the pseudovirus
94 neutralization assays. All single and double mutations showed 1.4- to over 28-fold increase in
95 resistance to Pemivibart. Some of these mutations significantly reduced SA55 neutralization; for
96 example, G404E and V503E were ~900-fold more resistant (Figure 1B). Furthermore, we
97 observed a strong positive correlation between fold-changes in Pemivibart neutralization and
98 hACE2 inhibition for these escape mutations, with the exception of K403N and N405D, which did
99 not affect viral infectivity (Figure S4 and S5B).
100
101 Finally, structural analyses were conducted to explore the mechanisms underlying resistance to
102 Pemivibart. These escape mutations can be classified into two categories based on their locations.
103 11 mutations emerged within the Pemivibart epitope, directly reducing or abrogating antibody
104 binding, and can be further divided into two groups (Figure S6). The first group consists of
105 mutations like G502I and T500N, which introduce steric clashes with both Pemivibart and the
106 ACE2 receptor (Figure S6A). The second, including mutations such as K403N and N405D,
107 disrupts interactions with Pemivibart without affecting ACE2 binding and thus allows JN.1 to
108 evade Pemivibart while retaining fitness (Figure S6B). Beyond these direct escape mechanisms,
109 16 NTD mutations outside the Pemivibart epitope are distributed into three clusters based on their
 bioRxiv preprint doi: https://doi.org/10.1101/2024.11.08.622746; this version posted November 10, 2024. The copyright holder for this preprint
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
available under aCC-BY-NC-ND 4.0 International license.

110 locations (Figure S7): (1) T236A, N234H/S, G232C, N196S, and Y200C/D at the NTD-RBD
111 interface; (2) K281E and K304E, located at the interface between the NTD and SD1 or SD2; and
112 (3) Mutations surrounding S31, including S31P, P39L, F55L, and R273K. Structural modeling
113 showed that the S31P mutation could alter the local conformation at the base of the NTD, similar
114 to the effect of the S31 deletion in KP.3.1.1 and F59S in XEC (Figure S7C). Our findings suggest
115 that these NTD mutations impart resistance by modulating spike conformation, enabling JN.1 to
116 evade Pemivibart neutralization through alternative mechanisms beyond simply disrupting virus-
117 antibody interactions within the epitope.
118
119 Discussion
120
121 In conclusion, Pemivibart retains broad activity against recent JN.1 sublineages but shows
122 substantially reduced potency against KP.3.1.1 and XEC. Therefore, monitoring the neutralizing
123 activity of Pemivibart is essential as JN.1 variants evolve. The viral escape profile of Pemivibart
124 provides valuable insights for predicting its clinical efficacy and guiding antibody optimization.
 bioRxiv preprint doi: https://doi.org/10.1101/2024.11.08.622746; this version posted November 10, 2024. The copyright holder for this preprint
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
available under aCC-BY-NC-ND 4.0 International license.

Figure 1
125 Figure and ledgend

A C K403E/N, G404E, N405D,

Spike mutations on JN.1 IC50 (µg/mL) Fold change in IC50 relative to JN.1
R498K, T500N, Y501D,

V1104L
Variants Vero-E6 Vero-E6-TA Vero-E6 Vero-E6-TA
H146Q
Q183H
S31del

Q493E
Pemivibart G502I, V503A/E, H505R

R346T
F456L
T22N

F59S
Pemivibart SA55 Pemivibart SA55 Pemivibart SA55 Pemivibart SA55
D614G 0.015 0.017 0.018 0.024 0.1 5.4 0.1 2.9
JN.1 0.121 0.003 0.191 0.009 1.0 1.0 1.0 1.0 T415I
T236A
KP.2 0.102 0.002 0.142 0.007 0.8 0.7 0.7 0.9 K440E/T
KP.2.3 0.651 0.005 0.653 0.023 5.4 1.6 3.4 2.7 G232C F375L
N234H/S
LB.1 0.427 0.007 0.753 0.013 3.5 2.3 3.9 1.6 RBD F371L
KP.3 0.474 0.003 0.778 0.010 3.9 1.1 4.1 1.1 Y200C/D
Y396H
KP.3.1.1 1.745 0.006 4.265 0.014 14.0 1.7 22.0 1.6
D142G F377L
XEC 1.204 0.005 4.247 0.021 10.0 1.6 22.0 2.5
V395F
JN.1-T22N 0.105 0.003 0.130 0.009 0.9 1.0 0.7 1.0 NTD D428Y
JN.1-S31del 0.547 0.007 0.661 0.020 4.5 2.3 3.5 2.3 N196S/T
JN.1-F59S 0.499 0.005 0.756 0.019 4.1 1.7 4.0 2.2 P384L
P39L SD1
JN.1-H146Q 0.212 0.007 0.171 0.013 1.8 2.1 0.9 1.5 F55L V332F/I
JN.1-Q183H 0.120 0.003 0.183 0.013 1.0 1.0 1.0 1.5 K281E T547I
JN.1-R346T 0.076 0.003 0.081 0.007 0.6 0.8 0.4 0.8 S31P R273K
JN.1-F456L 0.198 0.006 0.470 0.014 1.6 1.8 2.5 1.7 SD2
JN.1-Q493E 0.519 0.003 0.650 0.008 4.3 0.9 3.4 0.9
K304E
IC50 (µg/mL) <0.01 0.01-0.1 0.1-1 1-10 >10 Low High

B NTD mutations RBD mutations Double mutations

>67 1.0 >28 D142G/Y501D

>67 3.6 >28 P521S/A522G

>67 0.8 >28 R273K/Y501D

>67 0.4 >28 P384L/Y501D
8.4 1.7 >28 K304E/T547I
8.0 >28 S31P/K440T
Fold change
in IC50
2.9 0.8 1.4 G232C
28 N234D
17 N234H

1.3 1.1 >28 N405D

5.1 H505R
7.5 >945 >28 G404E
>67 6.8 >28 N196S

>67 10 >28 N234S

2.7 >28 D428Y

>67 1.0 2.6 R498K

2.8 >28 H519Y

11 Y200C
5.3 >28 Y200D

3.3 0.8 3.5 Y396H

1.0 0.8 >28 K403N

>67 2.9 >28 Y501D

7.1 1.8 2.8 N196T

4.8 1.9 9.8 K281E

7.6 >28 K304E

2.3 1.1 >28 K403E

16 >28 K440E

6.2 5.8 >28 V503A

8.9 842 >28 V503E
3.6 1.2 >28 T500N
>67 58 >28 V332F

>67 3.2 >28 V395F

>67 7.2 >28 T236A

>67 9.9 >28 F371L

4.4 >28 F375L
4.0 1.1 8.7 F377L

>67 1.0 >28 G502I

4.0 1.3 3.0 V332I

4.5 1.2 3.4 T415I

4.2 >28 P39L
2.7 >28 F55L

compared
1.0 1.0 1.0 JN.1

with JN.1
Pemivibart
2.5

>67 8.1
7.8 1.6

SA55

>67 55
hACE2
20
12

16
56

25

63

17
22

21

19
Resistance Sensitization

126
127 Figure 1. Neutralizing activity of Pemivibart against recent JN.1 sublieages and its escape
128 mutants.
129 A. Neutralizing activity of Pemivibart and SA55 against JN.1 sublineages in Vero-E6 and Vero-
130 E6-TMPRSS2-T2A-hACE2 (Vero-E6-TA) cells. Spike mutations of the indicated variants
131 relative to JN.1 are displayed in blue. The neutralization IC50 values and fold changes in IC50
132 compared to JN.1 for each variant are shown.
133 B. Neutralization resistance of Pemivibart escape mutants to Pemivibart, SA55 and hACE2. Data
134 are shown as fold changes in neutralization IC50 values relative to JN.1, with red indicating
135 resistance while blue indicating sensitization.
136 C. Location of pemivibart escape mutations in the SARS-CoV-2 spike protein. Escape mutations
137 within the Pemivibart epitope are highlighted in red font.
 bioRxiv preprint doi: https://doi.org/10.1101/2024.11.08.622746; this version posted November 10, 2024. The copyright holder for this preprint
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
available under aCC-BY-NC-ND 4.0 International license.

138 References:
139
140 1. FDA. FDA Roundup: March 22, 2024. 03/22/2024 2024. https://www.fda.gov/news-events/press-
141 announcements/fda-roundup-march-22-2024 (accessed 07/25/2024.
142 2. Bock A. New Guidance Helps Clinicians Use Pemivibart to Protect Immunocompromised Patients From
143 COVID-19. JAMA 2024.
144 3. Wang Q, Guo Y, Ho J, Ho DD. Pemivibart is less active against recent SARS-CoV-2 JN.1 sublineages. bioRxiv
145 2024: 2024.08.12.607496.
146 4. Planas D, Staropoli I, Planchais C, et al. Escape of SARS-CoV-2 Variants KP.1.1, LB.1, and KP3.3 From
147 Approved Monoclonal Antibodies. Pathog Immun 2024; 10(1): 1-11.
148 5. Yisimayi A, Song W, Wang J, et al. Repeated Omicron exposures override ancestral SARS-CoV-2 immune
149 imprinting. Nature 2024; 625(7993): 148-56.
150