In vivo
imaging of retrovirus infection reveals a role for Siglec-1/CD169 in multiple routes
1
of transmission
2
Kelsey A. Haugh
1
, Mark S. Ladinsky
2
, Irfan Ullah
3
, Ruoxi Pi
1
, Alexandre Gilardet
1
, Priti Kumar
3
,
3
Pamela J. Bjorkman
2
, Walther Mothes
1
, and Pradeep D. Uchil
1
4
1
Department of Microbial Pathogenesis, Yale University School of Medicine, New Haven CT
5
06511
6
2
Division of Biology and Biological Engineering, California Institute of Technology, Pasadena,
7
CA 91125
8
3
Department of Internal Medicine, Section of Infectious Diseases, Yale University School of
9
Medicine, New Haven, CT 06520
10
For correspondence:
11
Pradeep Uchil (pradeep.uchil@yale.edu)
and Walther Mothes
(walther.mothes@yale.edu)
12
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprint
this version posted October 20, 2020.
;
https://doi.org/10.1101/2020.10.20.347427
doi:
bioRxiv preprint
Abstract
13
Early events in retrovirus transmission are determined by interactions between incoming viruses
14
and frontline cells near entry sites. Despite their importance for retroviral pathogenesis, very little
15
is known about these events. We developed a bioluminesce
nce
imaging (BLI)-guided multiscale
16
imaging approach to study these events
in vivo
. Engineered murine leukemia reporter viruses
17
allowed us to monitor individual stages of
the
retrovirus life cycle including virus particle flow, virus
18
entry into cells, infection and spread for retroorbital, subcutaneous and oral routes. BLI permitted
19
temporal tracking of orally administered retroviruses along the gastrointestinal tract as they
20
traversed the lumen through Peyer’s Patch to reach the draining mesenteric sac. Im
portantly,
21
capture and acquisition of lymph-, blood-
and milk
-borne retroviruses spanning three routes, was
22
promoted by a common host factor, the
I-type lectin CD169, expressed on sentinel macrophages.
23
These results highlight how retroviruses co-opt the immune surveillance function of tissue resident
24
sentinel macrophages for establishing infection.
25
26
Keywords:
Retrovirus transmission, oral route, bioluminescence imaging, CD169
27
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprint
this version posted October 20, 2020.
;
https://doi.org/10.1101/2020.10.20.347427
doi:
bioRxiv preprint
Introduction
28
Retroviruses cause cancer and immunodeficiencies [1].
Once retroviruses establish viral
29
reservoirs,
it is difficult to eliminate infection as retroviral genomes are permanently integrated
30
into host DNA. Despite the clinical relevance of early processes by which incoming retroviruses
31
establish infection by navigating complex host tissue architecture
en route
to their first targets,
32
little is known about these events
[2-
4].
Retroviruses
like the human immunodeficiency virus
(HIV
-
33
1) can enter through the vaginal and rectal mucosa during sexual transmission, orally via milk
34
during mother
-to-child transmission, subcutaneously and intravenously through needle stick
35
injections during drug use and blood transfusions
[2, 5].
The majority of the Murine leukemia
36
viruses
(MLV) transmission in mice occur vertically from dam
-to-pup via ingestion of virus
-
37
containing milk through the gastrointestinal tract. MLV transmission can also occur parenterally
38
between male mice during infighting and via the
venereal route between infected male and female
39
mice
[6, 7]. Entry via different routes requires retroviruses to navigate diverse host tissue
40
architecture and overcome barriers for successful infection [8
-10].
Whether retroviruses exploit
41
common
host factors across these
transmission routes remain to be clarified.
42
We have previously used MLV as a model retrovirus to understand how retroviruses
43
establish infection in mice through the lymph or blood following subcutaneous and intravenous
44
delivery r
espectively
[11, 12]. We
found that sentinel macrophages lining blood/lymph-tissue
45
interfaces such as the subcapsular sinus in lymph nodes or the marginal zones in the spleen
46
function to filter out incoming retroviruses from circulation [11, 12]. The “fly
-paper”
like activity of
47
sentinel macrophages have been observed for various incoming
viruses and pathogens
[13-15]
.
48
The frontline position
of sentinel macrophages allows them to orchestrate downstream innate,
49
cell
-mediated, and humoral immune responses to incoming pathogens in the lymph and blood
[8,
50
12, 16-18].
These
macrophages naturally express the I
-type lectin Siglec
-1/CD169 that
51
specifically interact with sialosides present on retroviral envelopes [11, 19, 20].
CD169 expression
52
allows
sentinel macrophages to capture retroviruses and limits
their dissemination
[12]. However,
53
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprint
this version posted October 20, 2020.
;
https://doi.org/10.1101/2020.10.20.347427
doi:
bioRxiv preprint
retroviruses like MLV and HIV
-1 exploit their CD169 to promote infection of target lymphocytes
54
that sample antigens captured by sentinel macrophages
[11, 12, 21].
Whether the
observ
ed roles
55
for CD169
+
macrophages following subcutaneous and intravenous delivery, are of any relevance
56
for natural mother
-to-offspring transmission when viruses enter via
the gastrointestinal (GI) tract,
57
has remained unknown.
58
Given the complexity, functional diversity and length of the entire GI tract that can measure
59
over >7 cm even in neonatal mice, identifying portals of entry such as Peyer’s Patch (PP) in
60
underdeveloped intestines can be very challenging [22].
Here, we developed a
whole-body
61
bioluminescence imaging (BLI)
-based approach
to illuminate
areas where MLV concentrates to
62
traverse into the gut tissue from the lumen. We implemented BLI by developing a series of MLV
-
63
based reporter viruses to enable observation of specific stages of the retrovirus l
ifecycle
in vivo
,
64
including viral particle flow, entry into cytoplasm, first infection events, and spread. We first
65
validated this system by testing its ability to uncover new insights into previously studied
66
subcutaneous and intravenous transmission routes. Second, BLI-i
maging permitted temporal
67
tracking of various steps of virus infection for orally administered retroviruses along the
68
gastrointestinal tract as they traversed the lumen through the PP to reach the draining mesenteric
69
sac. Finally, we show that capture and acquisition of lymph-, blood-
and milk
-borne retroviruses
70
spanning three routes, was promoted by
a common host factor, CD169 expressed on sentinel
71
macrophages. Our results highlight how retroviruses co-opt the immune surveillance function of
72
tissue resident sentinel macrophages for establishing infection. Understanding these events will
73
inform
design improved prophylactic strategies that target prevention of virus acquisition and
74
establishment of infection.
75
76
Results
77
78
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprint
this version posted October 20, 2020.
;
https://doi.org/10.1101/2020.10.20.347427
doi:
bioRxiv preprint
Generation of reporter viruses to enable
visualization of
individual stages of retrovirus
79
infection
in vivo
80
We established a BLI-directed approach for studying individual stages of retroviral infection
in
81
vivo
by strategically inserting reporters into unique genomic loci of MLV
(Figure 1A). T
o track virus
82
particle flow, we generated bioluminescent virus
particles by introducing nanoluciferase (Nluc)
83
into the proline-rich region (PRR) of the MLV envelope (Env). Nluc
-Env-tagged
virus particles
84
produced using
a tripartite plasmid system (encoding Nluc
-tagged MLV Env, MLV Gag-Pol and
85
MLV
-LTR) exhibited 200 times more luciferase activity per virus particle (0.2 RLU/virion)
86
compared to viruses generated using the full
-length MLV genome (0.001 RLU/virion) (Figure 1B)
.
87
To monitor virus fusion and entry into the cytoplasm of cells, we packaged firefly luciferase (Fluc)
88
within the virus capsid as a C
-terminus fusion of MLV Gag (MLV Gag-Fluc
Env
wt
) and exploited
89
the ATP
-dependence of Fluc for its activity that is restricted to
the
host cell cytoplasm
in vivo
. This
90
strategy ensured that Fluc activity was exhibited when
both detergent (Triton X
-100) and ATP
91
were present (Figure 1C)
. A Gag
-Fluc labeled virus
in which MLV envelope was replaced by the
92
fusion-defective SFFV gp55 envelope (MLVGag-
Fluc
Env
FD
) served as a negative control. To
93
monitor single-round virus transduction, we utilized a replication-defective virus generated by co-
94
transfecting a dual BLI and GFP reporter (pMIG
-Nluc
-IRES
-GFP) in conjunction with MLV Gag-
95
Pol and Env [23]. In addition, we generated red-shifted reporter viruses encoding Antares, which
96
is a bioluminescence resonance energy transfer (BRET) reporter that enables superior deep-
97
tissue sensitivity over Nluc
in vivo
[24].
Finally, to permit longitudinal monitoring of progressing
98
infection, we generated replication competent MLV reporter viruses by introducing a shortened
99
internal ribosome entry site (IRES), 6ATRi, to drive Nluc expression downstream of viral Env
[23,
100
25, 26]. This
strategy enabled bi
-cist
ronic expression of Env and Nluc in infected cells. Infectivity
101
measurements revealed that MLV 6ATRi Nluc was ~ 60
% as infectious as wild-type MLV (Figure
102
1D)
.
103
104
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprint
this version posted October 20, 2020.
;
https://doi.org/10.1101/2020.10.20.347427
doi:
bioRxiv preprint
BLI
-driven characterization of blood-borne retrovirus infection
105
We revisited the previously studied intravenous (i.v.) route of MLV infection at a whole
-body level
106
by applying these
new reporter viruses. We challenged mice retroorbitally (r.o.)
with MLV Env
-
107
Nluc reporter viruses. BLI-driven virus tracking immediately after challenge showed that MLV
108
rapidly reached both liver and spleen within 30 s after administration and accumulated
109
predominantly at the spleen after passing through the liver (2-
3 min) (Figure 2A, B; Video 1). This
110
was consistent with the spleen being the main blood-filtering organ in mice. Interestingly, the
111
decaying luminescence over time (>30 min) in the spleen was revived when Nluc substrate
112
(furimazine) was re-administered (Figure 2-figure supplement 1). This indicated that viruses
113
remain captured at the spleen and the exhaustion of substrate contributed to the decay in the
114
signal. We next investigated virus entry into the host cell cytoplasm by utilizing MLV Gag-Fluc
-
115
tagged viruses with wild-type (Env
WT
) or fusion-defective Env
(Env
FD
). Inocula were equalized by
116
measuring the luciferase activity (relative light units; RLU) in detergent-lysed viral preparations
117
(Figure 2C). Mice challenged with MLV (r.o.) were monitored at 3 min, 40 min, and 1 h post-
118
infection (hpi
) using BLI. In contrast to animals infected with fusion defective
MLV, we observed
119
Fluc signal emerging at the spleen (1
hpi
) in animals infected with reporter viruses carrying wild
120
type MLV envelope (Figure 2D, E; p=0.0061). Taken together, our data indicated
that
blood borne
121
MLV was filtered rapidly at the spleen within 2-3
minutes and
entered
the cytoplasm of cells by
122
60 minutes after capture.
123
We next infected mice with WT FrMLV or MLV 6ATRi
-Nluc reporter virus and
compared
124
infection levels at 7 dpi.
In vivo
infectivity of MLV 6ATRi
-Nluc virus was reduced in comparison to
125
WT FrMLV (Figure 2F). This was not unexpected based on the reduced released infectivity of
126
MLV 6ATRi
-Nluc
in vitro
(Figure 1D), and the known effect of genomic reporter insertions on
127
retrovi
rus fitness (17). To visualize the
first round of infected cells and virus spread at the whole
128
animal level, we challenged mice with single-round MLV reporter virus (pMIG
-Antares) as well as
129
replication-
competent reporter MLV, and monitored replication dynamics using BLI every 2-3 days
130
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprint
this version posted October 20, 2020.
;
https://doi.org/10.1101/2020.10.20.347427
doi:
bioRxiv preprint
over the course of two weeks (Figure 2G). In contrast to the decline of luminescent signal
131
observed with single-round Antares
-encoding MLV, luminescent signal in organs infected by MLV
132
6ATRi
-Nluc increased over time, indicative of fresh rounds of infection (Figure 2G, H). We
133
observed infection in auricular and inguinal LN in addition to the spleen (Figure 2G, H). The
134
gradual decline of luminescent signal in mice infected with MLV 6ATRi
-Nluc is consistent with the
135
immune control of MLV infection in C57BL/6J (B6) mice [11, 12, 21, 27, 28]. Taken together, these
136
results demonstrate the utility of our bioluminescent reporter viruses in monitoring particle flow,
137
capture, cytoplasmic entry, transduction, and subsequent virus spread following intravenous
138
infection of mice. The validation of this reporter system also set the stage for applications to other
139
infection routes
.
140
BLI
-driven characterization of lymph-borne retrovirus infection
141
Intrafootpad (i.f.p.) infection is widely used to study subcutaneous (s.c.) infection and model
142
antigen trafficking to draining lymph nodes via lymphatics [29, 30]. The draining lymph node for
143
i.f.p. infection is the popliteal lymph node (pLN)
[14, 15, 31]. Previous studies using two-photon
144
intravital microscopy indicated that incoming viruses accumulated at the subcapsular sinus of the
145
pLN within a few minutes following viral challenge [11, 14, 15, 32].
We revisited this well
-studied
146
route using BLI imaging of incoming virus particle flow from the administration site to the target
147
organ by infecting mice i.f.p. with MLV Env
-Nluc. Incoming viruses accumulated rapidly at the
148
pLN, with detectable signal occurring within 1 minute 30 seconds p.i. (Figure 3A, B; Video 2). This
149
was indicative of lymph flow
-mediated dissemination of MLV to pLN and
was consistent with
150
previous multiphoton microscopy studies
[11]. However, we observed that most of the incoming
151
virus particles localized to the injection site at the footpad (Figure 3A, B). Quantification of virus
152
parti
cle accumulation in the footpad and pLN, displayed as photon flux (photons/sec), revealed
153
that virus accumulation in the footpad was over 200-fold higher than that in the pLN. This
154
observation remained constant for the entire imaging window. Even when virus eventually
155
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprint
this version posted October 20, 2020.
;
https://doi.org/10.1101/2020.10.20.347427
doi:
bioRxiv preprint
accumulated in the pLN, plateauing at ~9 minutes
pi., the level of virus in the footpad remained
156
~150-fold above that of the pLN (Figure 3A, B). Thus, our imaging analyses showed that only a
157
small fraction of incoming viruses surpass tissue barriers including collagen fibrils, muscle tissue,
158
and antigen-
capturing cells at the footpad to reach the draining site.
159
We next explored the location where viruses fuse and gain access to the host cell
160
cytoplasm. We challenged mice with MLV Gag-
Fluc
Env
WT
or
MLV Gag
-Fluc Env
FD
via the i.f.p.
161
route and monitored these events by BLI. Bioluminescent signal in the mouse footpads infected
162
with MLV Gag-Fluc Env
FD
did not increase over time (Figure 3C, D).
In contrast, luciferase activity
163
in footpads challenged with MLV Gag-Fluc Env
wt
increased over time, indicating progressive virus
164
access to cell cytoplasm
in vivo
(Figure 3C, D). We observed that fusion-competent
viruses
165
gained access to the host cell cytoplasm in the footpad as early as 3 minutes post-challenge
166
(Figure 3C, D). Despite
accumulation of virions in the pLN within minutes following challenge
167
(Figure 3A, B), we were unable to detect Fluc activity in the first 40 minutes of the continuous
168
imaging time frame (Figure 3C, D). However, viruses had gained cytoplasmic access in the pLN
169
by 6 hpi
(Figure 3C, D). The delay was likely due to the mode of infection at the pLN, where MLV
170
was first captured by CD169 macrophages before
trans
-infection of permissive B-1a cells [21,
171
27].
172
Next, we infected mice subcutaneously (s.c.) in the footpads with MLV 6ATRi
-Nluc or WT
173
MLV, harvested pLN at 3, 5, 7, 9, or 11 dpi
, and assessed the number of infected cells in individual
174
pLNs by flow cytometry using antibodies to MLV GlycoGag. WT MLV exhibited an expected
175
infection profile
in vivo
, peaking at 5
dpi
and subsequently decreasing due to immune control
176
(Figure 3F). As in r.o. infection (Figure 2F), MLV 6ATRi
-Nluc replicated with slower infection
177
kinetics compared to those of WT MLV and peaked at 7-8 dpi
instead of 5 dpi
(Figure 3E, F). We
178
also confirmed luciferase activity resulting from single cell suspensions isolated from pLN isolated
179
from mice infected with MLV 6ATRi
-Nluc at different time points (Figure 3G). Luciferase activity
180
mirrored infection curves obtained by flow cytometric enumeration of infected cells using
181
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprint
this version posted October 20, 2020.
;
https://doi.org/10.1101/2020.10.20.347427
doi:
bioRxiv preprint
antibodies to MLV GlycoGag. In contrast to r.o. challenge, virus infection during s
.c. challenge
182
was restricted to the draining pLN and new infection events were not observed beyond the target
183
organ before elimination by mounting immune responses in resistant B6 mice (Figure 3E and H).
184
These data are consistent with our earlier work describing how capture by CD169
+
macrophages
185
at pLN limits systemic spread and initiates effective immune responses [12, 33].
186
CD169
+
m
acr
ophages contribute to virus capture in the footpad
187
We next characterized the cell types that capture MLV at the footpad by challenging mice with
188
MLV Gag
-GFP particles (i.f.p) (Figure 4A). Surprisingly, immunostaining of footpad cryosections
189
(15 min post challenge) revealed that in addition to CD11c
+
DCs, CD169
+
macrophages also
190
reside in the footpad and predominantly captured MLV (Figure 4A). In many cases, we observed
191
virus-capturing CD169
+
macrophages in c
lose contact with DCs, indicative of synaptic cell
-cell
192
contacts (Figure 4A). Electron tomography revealed the presence of virus
-laden macrophages
193
with viruses present in membrane invaginations, as well as tethered to plasma membranes,
194
suggestive of CD169-mediated capture (Figure 4A, figure 4-supplement 1, Video 3). We also
195
observed viruses within cell
-cell contacts between virus
-capturing macrophages and DCs (Figure
196
4B, C). Quantification of cell
-associated viruses in tomographic sections were consistent w
ith
197
immunostaining data (Figure 4A) and revealed that incoming viruses predominantly associated
198
with macrophages (Figure 4-
figure 4 supplement 1) and to a lesser extent with DCs.
199
200
BLI
-driven characterization of oral route of retrovirus transmission
201
We utilized the capacity of BLI to pinpoint events of interest in extensive organs like the
202
gastrointestinal tract, by characterizing individual steps in the less understood oral route of MLV
203
transmission. To study mother
-to-
offspring transmission, we infected a lactating dam in the
204
mammary glands (s.c.) with WT MLV
car
rying a co-packaged Fluc reporter driven from
the
viral
205
LTR and allowed the infection to establish itself for 6 days. BLI confirmed the presence of
206
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprint
this version posted October 20, 2020.
;
https://doi.org/10.1101/2020.10.20.347427
doi:
bioRxiv preprint
luminescent signal in infected
teats at 6 dpi
(Figure 5A).
Subsequent transfer of
neonatal mice (1-
207
3 days old) for fostering resulted in successful transmission of MLV as seen by luciferase-positivity
208
in the gastrointestinal regions at 2 days post-transfer (
dpt) (Figure 5A).
Electron tomography of
209
neonatal
stomach contents
at 3 dpt,
revealed cell
-free viruses in milk (Figure 5B), congruent with
210
studies of mother
-to-offspring transmission of cell free MMTV to suckling pups [34, 35]. Due to
211
the internal location
and convoluted nature of the GI tract, necropsy was required for revealing
212
anatomic details of the infected regions. BLI analyses of GI tracts from neonates fostered by
213
infected dams revealed that viruses had established infection in the PP and mesenteric sacs by
214
8 dpt
(Figure 5C).
215
To explore the dynamics of virus transit in the intestinal tract, we orally inoculated 3-
day
216
old mice with replication
-defectiv
e MLV Env
-Nluc particles
and analyzed their distribution at 12,
217
24, or 48 hpi
via BLI. We observed a sequential movement of virus particles from the stomach at
218
12 hpi
to a striking,
temporal
accumulation in PP and mesenteric sacs of the small intestine
219
(Figure 6A, B).
There was a significant and simultaneous increase in particle accumulation
220
between 24 to 48 hpi
at both PP and mesenteric sac (Figure 6B).
These results indicated
that
221
incoming particles from PP can reach the draining mesenteric sac via the lymph drainage without
222
first undergoing replication. There was less frequent accumulation of incoming particles in cecal
223
patches,
likely because they are downstream of PP in the direction of intestinal traffic. Our data
224
reveal that PP, followed by mesenteric sacs, were the earliest intestinal structures to accumulate
225
incoming MLV.
226
To determine the tissues sites where incoming viruses entered the cytoplasm of intestinal
227
target cells, we orally infected neonatal mice MLV Gag-Fluc Env
WT
or
MLV Gag
-Fluc Env
FD
, which
228
served as
a control.
Fluc signal was first observed in PP (Figure 6C
). We saw a significant
229
temporal increase in access to the cytoplasm of target cells of PPs by incoming MLV Gag-Fluc
230
Env
WT
viruses compared to the fusion-deficient
MLV Gag
-Fluc Env
FD
control viruses
(Figure 6D
).
231
However, Fluc activity at the mesenteric sac began to increase weakly over control parti
cles only
232
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprint
this version posted October 20, 2020.
;
https://doi.org/10.1101/2020.10.20.347427
doi:
bioRxiv preprint
by 48 hpi
(Figure 6C
, E), despite early particle accumulation
(Figure 6B). Delayed cytoplasmic
233
access was a recurring theme, suggestive of the antigen presenting cell
-mediated trans
-infection
234
process, as seen in pLN, also
effective in mesenteric LNs. Finally, the use of single round (MLV
235
Nluc
-IRES
-GF
P) and replication competent viruses (MLV 6ATRi
-Nluc) confirmed true virus
236
infection events at both PP and mesenteric sacs at 96
hpi
by BLI (Figure
6F, G). Infection levels,
237
measured by flux, were expectedly lower in PPs compared to the draining mesenteric sacs, which
238
accumulated more viruses than PP
(Figure 6B, G).
239
240
Incoming retroviruses enter Peyer’s Patches through M cells
241
Microfold (M) cells in PP function as portals of entry into underlying lymphoid follicles for
242
particulate antigens, bacteria, and viruses such as MMTV present in the lumen
[36-38]. M cells
243
have been previously implicated in MMTV infection during oral transmission using mouse models
244
that have significantly reduced levels of M cells [
39].
Thus, we asked whether incoming MLV
245
infiltrates PP through M cells. We infected 3-day
-old mice with MLV Env
-Nluc, sampled intestines
246
at 48 hpi
to
identify luciferase-positive PPs,
and processed them for electron tomography to
247
delineate possible infiltration mechanisms (Figure 7A).
MLV particles were observed within
248
endosomes inside
of M cells
(Figure 7B). These data supported a contribution of M cells and a
249
transcytosis
mo
del for retrovirus infiltration from the intestinal lumen into the follicle region of the
250
PP [40-42].
251
252
CD169 contributes to virus particle accumulation and establishment of infection
in the GI
253
tract
254
We next asked if viruses are captured after arriving in the PP. Visualizing virus particles (MLV
255
Gag
-GFP) at
the
single cell level by immunostaining of tissue sections was challenging after
256
particles had traveled for days in the GI tract before arriving in the PP follicle. We therefore
257
resorted to performing retrovirus challenge
in adult mice by surgical
ly ligating
a region in the small
258
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprint
this version posted October 20, 2020.
;
https://doi.org/10.1101/2020.10.20.347427
doi:
bioRxiv preprint
intestine that contained PP. The
ligated
loop allowed us to increase the local particle
259
concentration of
MLV Gag
-GFP under live settings
[43]. After allowing the virus to be taken up for
260
1 h, PP region of the ligated intestine was processed for immunostaining and microscopy
(Figure
261
8). In addition to particles that were free or adhered to the epithelial cells, we observed particles
262
invading the epithelial barrier presumably through M cells
and dispersed in the PP follicle
(Figure
263
8A).
CD169
+
macrophages were not previously reported in the PP. Surprisingly, we
observed
264
CD169
+
macrophages located in the serosal side of the PP that had captured MLV Gag
-GFP
265
(Figure 8B). These data suggested a possible role for CD169
+
macrophages in capturing and
266
promoting retrovirus infection in intestinal PPs. We explored this possibility further in neonatal
267
mice. In contrast to their serosal location in adult PPs, CD169 macrophages were more dispersed
268
within developing neonatal PPs (Figure 9A
). We next asked whether CD169 plays a functional
269
role in promoting retrovirus acquisition during oral transmission from mother to offspring. We first
270
orally inoculated B6 and CD169
-/-
mice with MLV Env
-Nluc to monitor virus particle transit through
271
the intestine. BLI at 48 hpi
revealed a significantly reduced distribution of virus particles
272
throughout the intestines and within PP (~5-fold reduction; p<0.0001) as well as mesenteric sacs
273
(~3-fold
reduction; p=0.0082) in CD169
-/-
mice compared to B6
mice (Figure 9B
-D). We then asked
274
whether CD169 promoted infection in the mesenteric sac. We infected neonatal B6 and CD169
-/-
275
mice orally with WT FrMLV and assessed infection levels in individual mesenteric sacs via flow
276
cytometry for viral GlycoGag protein expression at 5
dpi
. Our analyses showed that CD169
-/-
mice
277
displayed a 26-fold reduction in establishment of infection in the mesenteric sac compared to B6
278
mice, (p=0.0011) (Figure 9E
). These results revealed that CD169 is a novel host factor that
279
contributes to retroviral acquisition via the oral transmission route.
280
281
Discussion
282
We developed a whole
-body imaging-guided, top-down approach to study retrovirus
283
infection
in vivo.
We engineered a series of reporter viruses to monitor individual steps of vir
us
284
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprint
this version posted October 20, 2020.
;
https://doi.org/10.1101/2020.10.20.347427
doi:
bioRxiv preprint
infection including virus particle flow, entry into cytoplasm, infection of cells, replication, and
285
spread in the context of living
animals.
We first validated this system by successfully
286
demonstrating its ability to uncover new insights into well
-studi
ed subcutaneous and intravenous
287
transmission routes. Importantly, BLI
permitted temporal tracking of orally administered
288
retroviruses along the GI tract as they traversed the lumen through PP to reach the drai
ning
289
mesenteric sac. We have also shown that a common host factor expressed on sentinel
290
macrophages, CD169, promotes retrovirus capture and acquisition throughout three routes. Our
291
results highlighted how retroviruses co-opt the immune surveillance function of tissue resident
292
sentinel macrophages to establish infection.
293
Our visual approach to virus infection
in vivo
revealed several novel facets of host-virus
294
interplay.
295
Route-specific tissue barriers:
Each route of entry comprises a unique set of tissue-specific
296
barriers that viruses must overcome for successful infection of the host. Our particle flow analyses
297
for the retroorbi
tal route revealed that most viruses reach the spleen, where they establish
298
infection after briefly passing through the heart and liver. In contrast, the i.f.p. route presented a
299
much more challenging barrier for viruses to reach their target organ, the pLN. Virus accumulation
300
in the footpad was 150-200-fold higher than that in the pLN (Figure 3B
). We
frequently observed
301
clusters of incoming virus particles trapped within the dense, tightly packed matrix of footpad
302
muscle
and collagen fibrils. In addition, virus particles were taken up by macrophages and DCs
303
that presumably trap them in a non-productive pathway (
Figure 4A-C)
and initiate
immune
304
responses,
in contrast to B and T cells located at the pLN that serve as amplifying hosts for MLV.
305
Despite the high levels of particle capture and infection seen near the injection site at the footpad,
306
our previous studies showed that the establishment of the infection at the draining pLN is
307
independent of cells migrating from the footpad [12]. During oral transmission, enveloped
308
retroviruses must withstand harsh conditions such as low pH in the stomach, digestive enzymes,
309
and bile salts [9, 44]
that can solubilize viral membranes. While retroviruses can be destroyed at
310
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprint
this version posted October 20, 2020.
;
https://doi.org/10.1101/2020.10.20.347427
doi:
bioRxiv preprint
pH as low as 4 [45], MLV is stable until pH 3 [46]. Furthermore, mice are most susceptible to MLV
311
infection at day 3 post-partum [38], when stomach acid production is low
[47]. Moreover, milk may
312
also shield retrovi
ruses by neutralizing secreted acid in the neonatal stomach.
Our
analyses
313
comparing
the
flux of input MLV Env
-Nluc
viruses
(~1 x 10
7
flux/s)
with
that of signal seen at 48
314
h in PP and mSac
(~1-2 x 10
5
flux/s
), suggest
that ~50-100-fold fewer viruses
are able to surmount
315
the
oral and gastrointestinal barriers. This
barrier is expected to increase
significantly
with
age
316
due to
the
upsurge in
production of
virus-inactivating factors like bile salts and aci
d thus rendering
317
adult mice resistant to
oral MLV transmission
[47].
318
Differential kinetics of virus particle arrival and cytoplasmic entry:
Following i.v. infection,
319
viruses accumulated in the spleen within 3 minutes and entered the cytoplasm of host target cells
320
by 60 minutes post-challenge (Figure 2D, E)
. In footpads, cytoplasmic entry corresponded with
321
virus particle arrival (Figure 3, A
-D)
. In pLN, however, viruses took longer
(more than 40 min)
to
322
enter host cell cytoplasm after arrival, with pLN luminescence first observed
at 6 h post-challenge
323
(Figure 3C, D)
.
Similarly, we saw a weak but delayed cytoplasmic entry of viruses in the
324
mesenteric sac compared to PP despite near simultaneous arrival of virus in both the organs
325
(Figure
6D
, E).
The delayed cytoplasmic entry in the p
LN and mesenteric sacs is likely explained
326
by the mode of infection. We have previously documented that in the spleen and pLN, CD169
+
327
macrophages capture incoming viruses and promote infection of target lymphocytes by a process
328
called
trans
-infection [11
, 12, 21].
Sentinel macrophages are resilient to MLV infection at early
329
time points as viruses are held at a distance of ~41 nm from the cell surface, corresponding to
330
the length of the CD169 ectodomain, a distance too far for the MLV Env to engage the rec
eptor
331
[11]. Following capture, viruses must transit through surface-
associated membrane invaginations
332
of CD169
+
sentinel macrophages before presentation to target cells for
trans
-
infection, delaying
333
cytoplasmic entry. In contrast, in footpads, incoming virus may directly interact with DCs, in
334
addition to the handoff from virus
-captur
ing macrophages, as observed by our
335
immunohistochemistry and electron tomography studies
(Figure 4B, C). Congruent with our
336
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprint
this version posted October 20, 2020.
;
https://doi.org/10.1101/2020.10.20.347427
doi:
bioRxiv preprint
observations of rapid virus entry at the footpad, vir
us-capturing DCs may internalize virus and
337
become infected to perform their natural role of antigen presentation unlike, CD169
+
338
macrophages
, which cross-present viruses [12, 33]. A similar situation may occur within PPs,
339
where viruses, after transiting through M cells, have direct access to target cells in the underlying
340
follicle in addition to availing CD169
+
macrophage promoted infection. In contrast, free viruses
341
entering mesenteric
LN
undergo sentinel macrophage-mediated
trans
-infection, delaying
342
cytoplasmic access like those seen in the pLN.
343
Differential capture of retroviruses by CD169 in spleen and liver:
We observed that
344
retroorbitally
-administered viruses briefly passed through the liver before accumulating in the
345
spleen
(Figure 2A, B). The vastly reduced virus
-retention in the liver, despite the presence of
346
sinusoidal CD169
+
Kupffer cells, was surprising. However, these results could be explained by
347
several
-fold lower levels of CD169 expression in Kupffer cells compared to CD169
+
SCS
348
macrophages in the LN or marginal zone metallophilic macrophages in the spleen [48]. A similar
349
predominant capture of intravenously delivered exosomes in spleen compared to liver was also
350
observed previously [48]. Alternatively, specific tissue environments may govern the capacity of
351
lectins to bind incoming viruses
as the binding capacity of Siglecs is often regulated by
352
endogenous ligands [49, 50].
353
Novel facets of oral retroviral transmission:
While requiring necropsy to increase resolution
354
and sensitivity, BLI assisted the study of long organs such as the GI tract,
as it can illuminate
355
areas of interest to guide directed investigations and shed light on the sequential course of
356
infection
events. Particle flow analyses with replication-defective reporters demonstrated that it
357
takes 24-48 h for viruses to reach portals of entry (PP) and accumulate in the draining mesenteric
358
sac. Interestingly, prior replication in PP was not required for transiting to
the
draining LN (Figure
359
6A
-C).
Incoming particles were able to access the lymph flow for transit to the mesenteric sac.
360
This contrasts with vaginal infection of SIV, in which local replication was critical for further virus
361
dissemination
[2, 3, 51]. The use of Fluc
-tagged viruses revealed that entry into the cytoplasm
362
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprint
this version posted October 20, 2020.
;
https://doi.org/10.1101/2020.10.20.347427
doi:
bioRxiv preprint
occurs at ~24 hpi
in the PP and begins at ~48
hpi
in the mesenteric sac which
houses mesenteric
363
LN
s. This was followed by infection at 96 hpi
(Figure
6F
, G).
As with other infection routes we
364
tested, CD169 expression on sentinel macrophages was crucial for promoting oral transmission
365
(Figure
9).
Particle retention and subsequent infection at PP and mesenteric sacs were
366
significantly reduced in the absence of CD169. These data were reminiscent of our previous
367
study, where we observed a CD169-requirement to promote MLV infection at both pLN and
368
spleen
[11, 12].
Thus, our studies
revealed the existence of CD169-mediated capture and
369
infection promotion as a second crucial step downstream of likely entry through M cells that
370
augments oral retroviral transmission.
371
Overall, our BLI-guided analyses have highlighted how retroviruses, during a million years
372
of co-evolution, have co-
opted CD169, primarily used for immune surveillance by tissue resident
373
sentinel macrophages, as a common host factor for promoting host colonization via various
374
naturally
occurring
infection routes. Our study opens avenues for a localized CD169-blockade
375
based strategy to curb retrovirus acquisiti
on and transmission.
376
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprint
this version posted October 20, 2020.
;
https://doi.org/10.1101/2020.10.20.347427
doi:
bioRxiv preprint
Materials and Methods
377
378
Ethics statement
379
All animal experiments were performed according to protocols approved by the Institutional
380
Review Board and the Institutional Animal Care and Use Committee (IACUC) of Yale University.
381
Yale University is registered as a research facility with the United States Department of Agriculture
382
(USDA), License and Registration number: 16-R-0001. It also is fully accredited by the
383
Association for Assessment and Accreditation of Laboratory Animal Care (AAALA
C). An Animal
384
Welfare Assurance (#A3230-01) is on file with OLAW-NIH.
385
386
Mice
387
C57BL/6 (B6) and BALB/cJ, mice were obtained from the Jackson Laboratory (Bar Harbor, ME).
388
CD169
−/−
mice (B6 background) were from Paul Crocker
[52]
, University of Dundee, UK. Animals
389
were housed under specific pathogen-free conditions in the Yale Animal Resources Center
390
(YARC). The Institutional Animal Care and Use Committees (IACUC) and the Institutional
391
Biosafety Committee of Yale University approved all experiments. 6–8-week
-old male and female
392
mice were used for all experiments involving adult mice. Breeder mice acting as foster mothers
393
for litter transfer experiments were 3-6 months of age. Oral inoculation experiments in neonatal
394
mice were performed on 3-day
-old mice.
395
396
Gen
eration of viral vector plasmids
397
Virus-encoding plasmids were generated using Gibson Assembly (NEB Gibson Assembly kit,
398
NEB, Ipswich, MA). Insert amplicons
containing 25 bp overla
ps
to target regions
were generated
399
using Kapa HiFi Hotstart
TM
high-fidelity
polymerase (Kapa Biosystems/Sigma-Aldrich, St. Louis,
400
MO)
using a
touchdown PCR
protocol.
MLV Env
-Nluc: For generating full
-length MLV Env
-Nluc
401
construct,
MLV pLRB303 Env-PRR-
GFP
plasmid
encoding a full length MLV genome [53]
was
402
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprint
this version posted October 20, 2020.
;
https://doi.org/10.1101/2020.10.20.347427
doi:
bioRxiv preprint
digested
with NheI to release the
GFP
cloned into the
Proline R
ich
Region (PRR)
of MLV
403
envelope. Nluc was amplified from the pNL1.1 plasmid (Promega, Madison, WI)
with primers
to
404
ensure
in-frame Nluc expression as well as containing 25 bp overlap to the insertion site
to allow
405
cloning by
Gibson Assembly. A similar strategy was used to generate MLV envelope expressor
406
plasmid for insertion of Nluc
in-frame
into PPR region.
MLV Gag
-GFP
: MLV Gag
-GFP
in the full
407
length viral context
was described previously
[11].
MLV Gag
-Fluc Env
WT
: GFP was released
from
408
MLV Gag
-GFP
using EcoRI and HindIII digestion. Fluc was amplified
from
pGL4.32[luc2P NF-
409
kB
-RE]
plasmid (Promega, Madison, WI)
and inserted into the vector backbone using Gibson
410
Assembly. For MLV Gag
-Fluc
Env
FD
generation, MLV envelope was replaced with SFFV envelope
411
using restriction digestion and ligation. SFFV envelope was amplified from
pBR322-
SFFV LS, a
412
gift from Leonard Evans and Frank Malik (NIH).
Full
-length MLV 6ATRi
-Nl
uc: gBlocks
TM
encoding
413
6ATRi
-Nluc
along with flanking region
were obtained from Integrated DNA Technologies
414
(Coralville, IA). Amplified gBlocks and Full
-length Friend MLV plasmid were digested with ClaI
415
and
BlpI and assembled into the FrMLV backbone via Gibson Assembly.
pMIG
-Nluc
-IRES
-GFP:
416
pMIG
-Nluc
-IRES
-GFP was described previously [23]. pMIG
-Nluc
-IRES
-GFP
which was
417
described previously [23]. pMIG
-Fluc
-IRES
-mCherry
was a gift from Xiaoping Sun (Addgene
418
plasmid # 75020; http://n2t.net/addgene:75020; RRID:Addgene_75020)
. pMIG
-Antares:
pMIG
-
419
Antares was generated by replacing the IRES GFP cassette of pMIG
-w with Antares luciferase
420
from the pNCS
-Antares. pMIG
-w was a gift from Luk Parijs (Addgene plasmid # 12282;
421
http://n2t.net/addgene:12282; RRID:Addgene_12282)
[54], and
pNCS
-Antares was a gift from
422
Michael Lin (Addgene plasmid # 74279; http://n2t.net/addgene:74279; RRID:Addgene_74279)
423
[24]. Plasmids were transformed into DH5
ɑ Max Efficiency competent
E. coli
(ThermoFisher,
424
Waltham, M
A). E. coli were grown overnight in 1L cultures of 2X
Yeast extract Tryptone (YT)
425
media at 30
-37
°C under shaking conditions. Plasmids were isolated using Machery-Nagel DNA
426
preparation kits. Reporter gene expression was tested by transfecting 50 ng of plasmid and 450
427
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprint
this version posted October 20, 2020.
;
https://doi.org/10.1101/2020.10.20.347427
doi:
bioRxiv preprint
ng of pcDNA3.1 into HEK293 cells seeded in 24-well plates. 24 h post-transfection, reporter gene
428
expression was monitored via flow cytometry and/or luciferase assay.
429
430
Generation of viral stocks
431
Plasmids encoding each vector were transfected into HEK293 cells seeded in 10
cm plates using
432
Fugene 6
TM
(Promega, Madison, WI) or polyethyleneimine. 12 μg of plasmid encoding full
-length
433
viral vectors were transfected into each 10
cm plate. Replication-defective viruses were produced
434
via transfection of 12 μg total DNA consisting of a mixture of MLV Gag
-Pol, LTR
-Reporter plasmid,
435
and Ecotropic MLV Envelope in a 5:5:2 ratio. ViralBoost
TM
reagent (ALSTEM, Richmond, CA) was
436
added to producer cell plates 12
h following transfection. 48 h following transfection, culture
437
supernatants were collected and filtered through 0.45
μM low protein
-binding filters (Pall
438
Corporation, Port Washington, NY). Viral stocks were aliquoted into 2-mL tubes and stored at
-
439
80
°C.
440
441
Virus Titration
442
For titration of viral stocks, DFJ8 chicken fibroblasts (50,000 cells/well, 48-well plate) were
443
infected with varying dilutions of viral stocks in the presence of 8 μg/mL polybrene (Sigma
Aldrich).
444
48 h following infection, infected DFJ8 cells were analyzed by flow cytometry using antibodies to
445
MLV GlycoGag. For preparation of single-cell suspensions, infected cells were incubated with
446
Accutase
TM
(StemCell technologies, Vancouver, Canada) f
or 5 minutes at 37°C. Accutase was
447
neutralized via addition of RPMI containing 10
% FBS. Cells were centrifuged for 5 minutes at
448
~110 x g and resuspended in PBS containing 1
% BSA. Cells were fixed with 4
% PFA for 7
449
minutes at room temperature. PFA was neutralized by addition of 0.1 M glycine in PBS. Cells
450
were washed with PBS
-0.1M
glycine twice and resuspended in FACS buffer
solution (5% FBS, 1
451
% BSA, 0.2
% Gelatin
in PBS) for 15 minutes at room temperature. Cells were incubated wit
h
452
1:500 anti
-GlycoGag antibody conjugated with Alexa 647 at room temperature for 1h. Cells were
453
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprint
this version posted October 20, 2020.
;
https://doi.org/10.1101/2020.10.20.347427
doi:
bioRxiv preprint
washed twice with and resuspended in MACS Buffer (
1X PBS, 2
mM EDTA, 1
% BSA). Infectious
454
units of fluorescent protein-encoding constructs were additionally
enumerated
by estimating the
455
number of GFP or mCherry-expressing cells. Viral titers were determined by calculating infectious
456
units per mL, based on numbers of infected cells resulting from each volume of virus supernatant
457
used in titration.
458
459
In vitro
luciferase assays
460
Single cell suspensions obtained from tissues or infected cells were lysed in 1X Passive Lysis
461
Buffer (Promega)
for 5 min at 37°C
and lysates were added to white-bottom 96-well flat
-bottom
462
plates (Costar, Corning, NY).
Luciferase activity was measured after adding appropriate substrate
463
(Promega Firefly Luciferase Assay substrate for firefly luciferase, or Promega nanoGlo
TM
464
nanoLuciferase substrate for N
Luc and Antares, diluted 1:40 in PBS per manufacturer’s
465
instructions)
using a Berthold luminometer (Berthold technologies, Bad Wildbad, Germany).
466
Luciferase
activity associated with virions were performed on partially purified viruses that were
467
sedimented through a 15 % sucrose cushion in PBS at 25000
x g
for 2 h at 4 °C. S
edimented
468
viral pellets w
ere
resuspended in 0.1
% BSA/PBS
and
diluted accordingly for detection within the
469
luminometer linear range. For MLV Gag-Fluc virion luciferase assays, sedimented viral pellets
470
were resuspended in 0.1
%BSA/PBS or Passive Lysis Buffer (Promega). ATP
-free or ATP
-
471
containing (150 μM ATP) substrate solutions were prepared containing 150 μg/mL D
-Luciferin
472
(Goldbio, St. Louis, MO), 100
mM Tris pH 8, and 5 mM MgCl
2
. Relative light units (RLU) were
473
determined by taking luciferase readings of lysis buffer or PBS/0.1
% BSA.
474
475
Retrovirus administration
476
Virus stocks were stored at -80
°C, thawed at 37
°C, quickly placed on ice, and concentrated by
477
sedimentation through a 15 % sucrose-PBS cushion for 2 h at 4
°C, at 25,000 x g. After
478
sedimentation, virus pellets were resuspended
in endotoxin free 0.1
% BSA
in PBS
at appropriate
479
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprint
this version posted October 20, 2020.
;
https://doi.org/10.1101/2020.10.20.347427
doi:
bioRxiv preprint
luciferase light units or infectious units (IU) for administration. Retroorbital
(100 μL)
and
480
subcutaneous (intra footpad or intra mammary glands
; 25 μL
)
injections
were carried out using
481
an insulin syringe with 31 G needle
on anesthetized mice (
0.5 -
5 % isoflurane
delivered using
482
precision Dräger vaporizer with oxygen
flow rate
of 1 L/min).
During retroorbital and subcutaneous
483
virus inoculations
, mice received
1 x
10
5
IU
of MLV Env
-Nluc
for
viral particle flow monitoring, 2 x
484
10
4
IU of
MLV Gag
-Fluc
Env
WT
or MLV Gag-Fluc
Env
FD
, for
monitoring of cytoplasmic entry, or
5
485
x 10
5
IU
of MLV 6ATRi
-Nluc, MLV
-Antares or WT MLV
for monitoring of longitudinal virus spread
486
and infectivity comparison.
For mother-to-offspring transmission experiments, dams were
487
inoculated with
1 x 10
7
IU of MLV Fluc
-IRES
-mCherry
subcutaneously distributed
into
the
488
mammary gland 6 days prior to transfer of neonates
for fostering.
For oral inoculation of neonatal
489
mice
, appropriate amounts
of virus were
resuspended in
15 μL sterile
endotoxin free PBS
490
containing 0.1
% BSA and 5
% sucrose and fed using a p10 pipette tip.
Mice
were
orally
491
inoculated
with
1 x 10
6
IU of MLV Env
-Nluc
for monitoring virus particle flow, 2 x 10
4
RLU of MLV
492
Gag
-Fluc for monitoring virus entry into cytoplasm, or 1 x
10
6
IU of
MLV 6ATRi
-Nluc or MLV pMIG
-
493
Nluc for monitoring establishment of infection.
494
495
Bioluminescence Imaging
(BLI)
496
Image acquisition:
All m
ice were anesthetized via isoflurane inhalation (3 – 5
% isoflurane, oxygen
497
flow rate of 1.5 L/min)
prior
and during
BLI using the XGI-8 Gas Anesthesia System. Images were
498
acquired
with an IVIS Spectrum
®
(PerkinElmer) and analyzed with the manufacturer’s Living
499
Image v
4.7.3
in vivo
software package.
Image a
cquisition exposures were set to auto, with
500
imaging parameter preferences set in order of exposure time, binning, and f/stop, respectively.
501
Images were acquired with luminescent f/stop of 1, photographic f/stop of 8. Binning was set to
502
medium.
503
Short interval i
maging
for
particle flow:
Prior to imaging MLV Env
-Nluc virion flow, mice received
504
furimazine (Promega) diluted 1:40 in PBS. 100 μL
diluted
furimazine were administered
505
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprint
this version posted October 20, 2020.
;
https://doi.org/10.1101/2020.10.20.347427
doi:
bioRxiv preprint
retroorbitally. Image sequences were then acquired at 30 sec
intervals
following administration of
506
MLV Env
-Nluc
(r.o or
i.f.p).
Image sequences
were assembled and converted to videos using
507
Image J
.
508
Imaging of adult mice:
Mice infected with Antares
- or
Nluc
-encoding viruses
received 100 μL
509
furimazine (NanoGlo
TM
furimazine, Promega, Madison, WI) diluted 1:40 in sterile
endotoxin-free
510
PBS (r.o.
) before imaging. Mice infected with Fluc
-carrying viruses received in vivo grade D
-
511
Luciferin (15 m
g/mL in sterile endotoxin-free PBS, GoldBio) (r.o.)
before imaging.
512
Imaging of neonatal mice:
513
For non-invasive imaging of neonatal mice that were
fed from infected dam
s, 25 ul of pre-warmed
514
in vivo
grade D-luc
iferin
in PBS (15 mg/mL)
was injected subcutaneously in the scruff of the neck.
515
The substrate was allowed to diffuse for 10 min before
imaging the mice using IVIS. For oral virus
516
administration, neonatal mice were orally inoculated with various reporter viruses suspended
in
517
15 uL of 5
% sucrose in PBS. Luciferase-specific substrate was subcutaneously injected as
518
above,
10 min before
euthansia.
Infected areas of interest were identified by carrying out
whole-
519
body imaging following necropsy [23]. Infected
regions
indicated by luminescent signal were
520
sampled and washed in PBS to remove residual blood and placed onto a glass plate. Additional
521
droplets of furimazine in PBS (1:40) or D
-Luciferin (15 mg/mL) were added to organs and soaked
522
in substrate for 1-2 min before BLI
for quantitation.
523
Image Processing:
Comparative images were compiled and batch-processed using the image
524
browser with collective luminescent scales.
Photon flux was measured as luminescent radiance
525
(p/sec/cm
2
/sr). During luminescent threshold selection for image display, luminescent signals
526
were regarded as background when minimum threshold levels resulted in displayed radiance
527
above non-tissue-containing or known uninfected regions.
528
529
Single cell preparation from mouse tissue
530
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprint
this version posted October 20, 2020.
;
https://doi.org/10.1101/2020.10.20.347427
doi:
bioRxiv preprint
Popliteal lymph nodes, mesenteric sacs
and spleens harvested after necropsy were disrupted in
531
serum free media, treated with Liberase TL (0.2 mg/ml, Sigma
-Aldrich) and DNase I (20 mg/ml,
532
Roche) at 37
°C
for 20 min and passed through a 70 μ
m cell strainer (Falcon, Cat # 352350).
533
Splenic cell suspensions were treated additionally with red blood cell lysis buffer at room
534
temperature for 10 min (Sigma-Aldrichor BioLegend
Inc
) for removing RBCs to obtain single cell
535
suspensions. Single cells suspensions from each lymphoid tissue were fixed with 4 %
536
paraformaldehyde (
PFA,
Electron Microscopy Sciences) before processing for flow cytometric
537
analysis.
538
539
Flow cytom
etry
540
Fixed
Single
-cell
suspensions were washed twice with 0.1 M glycine in PBS
to neutralize excess
541
PFA. Cells were blocked
with
CD16/32 antibodies (Biolegend® Inc, San Diego, CA
) in
FACS
542
staining buffer
(5% FBS, 1% BSA, 0.2% Gelatin) for 15 minutes
to 1 h. Cells were then stained
543
with
Alexa647 conjugated antibodies to Glycogag
(mAb34 hybridoma)
incubated with antibodies
544
for 1-2h at room temperature. For intracellular staining cells were permeabilized with staining
545
buffer supplemented with 0.2 % Triton X
-100.
Flow cytometry was performed on a Becton
546
Dickinson Accuri
TM
C6 benchtop cytometer. Data were analyzed with Accuri C6 or FlowJo v10
547
software (Treestar, Ashland OR).200,000 –
500,000 viable cells were acquired for each sample.
548
549
Retrovirus challenge by Ileal ligation
550
Ileal ligation
was
performed on anesthetized mice. Mice were first anesthetized with a
551
ketamine/xylazine cocktail (ketamine 15 mg/mL xylazine 1mg/mL) at 0.01 mL per g of body
552
weight. Mice were then placed on isoflurane inhalation anesthesia using a Dräger vaporizer (1.5-
553
2% isoflurane, flow rate 1L/min). A small, ~5-mm incision was made above the abdominal cavity
554
to expose the peritoneal muscle, in which a small ~3-mm incision was made above the intestine.
555
The small intestine was carefully extracted using forceps. Small intestinal regions of interest were
556
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprint
this version posted October 20, 2020.
;
https://doi.org/10.1101/2020.10.20.347427
doi:
bioRxiv preprint
ligated by tying surgical suture (undyed, Vicryl braided P
-3 polyglactin-coated J494G suture,
557
Ethicon, Somerville, NJ) gently around serosal intestinal walls surrounding Peyer’s Patches
. The
558
suture needle was inserted through the mesenteric membrane,
with care to avoid blood vessel
559
obstruction. Knots were gently tied at each end of the Peyer’s Patch, spanning a ~1-2-
cm length.
560
MLV Gag
-GFP v
iruses
corresponding to 1-4 x 10
5
IU based on comparative western blot analyses
561
with antibodies to Gag with equivalent amounts of WT FrMLV. were administered into the
562
intestinal lumen through the intestinal wall using a 31-G needle. Total volume did not exceed 50
563
μL. Following the intes
tinal loop, excess suture was trimmed from knots and the intestine was
564
carefully threaded back through the incision. The incision was sealed very gently using a small,
565
low
-tension binder clip, while the mouse remained under anesthesia until euthanasia at the end
566
of the terminal surgical procedure. Intestinal loop inoculations did not exceed 1 h. Anesthesia was
567
closely monitored during the duration of the surgery and intestinal loop inoculation. Anesthetic
568
planes were monitored every 15 minutes or more
frequently
based on
heart rate, breathing rate
569
and depth, and noxious stimuli reflexes as recommended by the Yale IACUC. At the end of 1 h,
570
the mouse was sacrificed,
and the ligated intestinal region processed for cryosectioning and
571
immunostaining.
572
573
Cryo
-immunohistology
574
Non
-fluorescent-protein
-containing tissue samples were harvested at indicated time points and
575
fixed in 1X PBS containing freshly prepared 4 % PFA for 1 h at 4°C. Fluorescent-protein-
576
containing samples were harvested and fixed in
periodate-lysine
-paraformaldehyde (PLP) fixative
577
(1X PBS containing 1% PFA, 0.01 M sodium m
-periodate and 0.075 M L-Lysine) for 30 min to 1
578
h to preserve fluorescent protein fluorophores. Tissue samples were washed with PBS,
579
dehydrated in a sucrose gradient consisting of 1h incubation at room temperature in 10, 20, or
580
30% sucrose in PBS, embedded and snap-frozen in Tissue-Tek® O.C.T.
TM
compound (Sakura
581
Finetek, Torrance, CA) and stored at –
80°C. 15
μm tissue sections were cut on a Leica cryostat
582
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprint
this version posted October 20, 2020.
;
https://doi.org/10.1101/2020.10.20.347427
doi:
bioRxiv preprint
at -20°C and
placed onto Superfrost Plus
TM
slides (Thermo, Waltham, MA). Tissue sections were
583
dried at 37°C for 15 minutes and stored at -20 for later use or
rinsed in PBS for staining. Slides
584
were washed in PBS (for stains of cell
-surface proteins) or permeabilized w
ith PBS containing
585
0.2% Triton X
-100 (for stains of intracellular proteins) and treated with Fc receptor blocker
586
(Innovex Biosciences), Richmond, CA or Staining
solution (5% FBS, 1% BSA, 0.2% Gelatin)
587
before staining with indicated antibodies in PBS containing 2 % BSA. Staining was performed
588
with following
antibodies:
CD169
-AF
647, CD169-AF
594,
CD11c
-eFluor450, CD11c
-AF
594,
589
CD11c
-AF
647,
CD68
-AF594 were from Biolegend® Inc, San Diego, CA;
GlycoGag
-CF
647
590
(mAb34 hybridoma, house-conjugated([12, 55].
Stained sections were washed with PBS
and a
591
final rinse with water to minimize salt
precipitation
and
mounted using ProLong® Glass antifade
592
reagent (
Invitrogen,
ThermoFisher) and Fisher
Finest thick coverslips (ThermoFisher). Mounted
593
slides were sealed with clear nail polish and cured for 1
h or overnight at 37°C. Slides were
594
analyzed by confocal microscopy using a Leica TCS SP8 microscope equipped with a white light
595
and argon laser, and a Nikon W-1 Spinning Disk microscope. The images were processed using
596
Volocity™ version 6.3 software (PerkinElmer, Waltham, MA, USA) and Nikon Elements software
597
(Nikon,
Tokyo, Japan). Figures were assembled with Photoshop CC and Illustrator CC (Adobe
598
Systems, San Jose, CA, USA).
599
600
Electron Microscopy and Dual-Axis Tomography
Following removal or disarticulation, tissues
601
were lightly fixed with 3
% glutaraldehyde, 1
% paraformaldehyde, 5
% sucrose in 0.1M sodium
602
cacodylate trihydrate to render them safe from virus infectivity. Tissues were further dissected in
603
cacodylate buffer, rinsed with cacodylate containing 10% Ficoll (70kD, Sigma) which served as
604
an extracellular cryoprotectant, placed in brass planchettes (Tell Pella, Inc, Redding, WA) and
605
ultra-rapidly frozen with a HPM-010 high pressure freezing machine (BalTec/ABRA, S
witzerland).
606
Samples were then transferred under liquid nitrogen to cryo-tubes (Nunc) containing a frozen
607
solution of 2
% osmium tetroxide, 0.05
% uranyl acetate in acetone and placed in an
AFS
-2 freeze-
608
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprint
this version posted October 20, 2020.
;
https://doi.org/10.1101/2020.10.20.347427
doi:
bioRxiv preprint
substitution machine (Leica Microsystems,
Wetzlar, Germany
). Tissues were freeze
-substituted
609
at
-90
°C for 72 h, warmed to -20 °C
over 12 h, and held at that temperature for an additional 12
610
h before warming to room temperature and infiltrating into Epon-Araldite resin (Electron
611
Microscopy Sciences, Port Washington PA). Samples were flat-embedded between two Teflon-
612
coated glass microscope slides and the resin polymerized at 60
°C for 48 h.
Embedded
tissue
613
blocks were observed by light microscopy to ascertain preservation quality and to select gross
614
regions of
interest. Blocks were extracted with a scalpel and glued to plastic sectioning stubs prior
615
to sectioning. Semi
-thick (300-400 nm) sections were cut with a UC6 ultramicrotome (Leica
616
Microsystems) using a diamond knife (Diatome, Ltd., Nidau,Switzerland) Sec
tions were placed
617
on formvar
-coated copper
-rhodium slot grids (Electron Microscopy Sciences) and stained with 3
618
% uranyl acetate and lead citrate. Colloidal gold particles (10 nm) were placed on both surfaces
619
of the grids to serve as fiducial markers for tomographic image alignment. Grids were placed in a
620
dual
-axis tomography holder (Model 2010, E.A. Fischione Instruments, Export PA) and imaged
621
with a Tecnai TF30ST-FEG transmission electron microscope (300 KeV; ThermoFisher
622
Scientific). Images were recorded
with a 2k x 2k CCD camera (XP1000; Gatan, Pleasonton, CA).
623
Tomographic tilt-series and large-area montages were acquired automatically using the SerialEM
624
software package
[56].
For dual
-axis tomography, images were collected at 1° intervals as
625
samples were tilted +/-
64°. The grid was then rotated 90° and a second tilt-
series was acquired
626
about the orthogonal axis Tomograms were calculated, analyzed and modeled using the IMOD
627
software package [57, 58]
on MacPro and iMac Pro computers (Apple, Inc, Cupertino, CA
). Cell
628
types and frequency within tissue sections were identified using 2D montaged overviews. Virus
629
particles and infected cells were further characterized in 3D by high-resolution electron
630
tomography.
631
632
Statistical Analyses
633
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprint
this version posted October 20, 2020.
;
https://doi.org/10.1101/2020.10.20.347427
doi:
bioRxiv preprint
Statistical
comparisons were performed using non-parametric Mann-Whitney tests (two-tailed)
634
available in GraphPad Prism software (Graph Pad Software, La Jolla, CA, USA). A difference was
635
considered significant if P < 0.05.
636
637
Acknowledgments
638
This work was supported by NIH grants R01 CA098727 to W.M. and P50GM082545 to W.M. and
639
P.J.B.;
R33AI122384 and R01AI145164 to P. K.,
the Flow Cytometry Shared Resource of the
640
Yale Cancer Center P30 CA016359, Yale Center for Cellular and Molecular Imaging S10
641
OD020142, and a fellowship from the China Scholarship Council –
Yale World Scholars to R.P
.
642
We thank the Kavli Nanoscience Institute at Caltech for maintenance of the TF
-30 electron
643
microscope.
644
645
Author Contributions
646
K.A.H. carried out all the BLI imaging. KA.H
and P.D.U. contributed to executing experiments,
647
data analysis, and figure generation. K
.A.H. P.D.U., and W.M. shared the conceptualization,
648
experimental design, interpretation, and manuscript preparation. M.S.L. carried out electron
649
tomography of tissue samples. R.P.
assisted in carrying out experiments, primer design and
650
maintaining mouse
colon
ies
. A.G.
assisted
with cloning and generation of reporter constructs.
651
P.J.B.
and P.K.
contributed to interpretation and discussion of the work.
652
653
Competing interests
654
The authors declare no competing interests.
655
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprint
this version posted October 20, 2020.
;
https://doi.org/10.1101/2020.10.20.347427
doi:
bioRxiv preprint
References:
656
1.
Blattner WA. Human retroviruses: their role in cancer. Proc Assoc Am Physicians.
657
1999;111(6):563-72. doi: 10.1046/j.1525-1381.1999.99210.x. PubMed PMID: 10591085.
658
2.
Haase AT. Targeting early infection to prevent HIV
-1 mucosal transmission. Nature.
659
2010;464(7286):217-23. doi: 10.1038/nature08757. PubMed PMID: 20220840.
660
3.
Haase AT. Early events in sexual transmission of HIV and SIV and opportunities for
661
interventions. Annu Rev Med. 2011;62:127-39. doi: 10.1146/annurev
-med
-080709-124959.
662
PubMed PMID: 21054171.
663
4.
Haase AT. Overview of the landscape of HIV prevention. Am J Reprod Immunol.
664
2014;71(6):490-4. doi: 10.1111/aji.12228. PubMed PMID: 24702688.
665
5.
Friedland GH, Klein RS. Transmission of the human immunodeficiency virus. N Engl J
666
Med. 1987;317(18):1125-35. doi: 10.1056/NEJM198710293171806. PubMed PMID: 3309656.
667
6.
Portis JL, McAtee FJ, Hayes SF. Horizontal transmission of murine retroviruses. J Virol.
668
1987;61(4):1037-44. PubMed PMID: 3029398; PubMed Central PMCID: PMC254060.
669
7.
Buffett RF, Grace JT, Jr., DiBerardino LA, Mirand EA. Vertical transmission of murine
670
leukemia virus through successive generations. Cancer Res. 1969;29(3):596-602. PubMed
671
PMID: 4304387.
672
8.
Uchil PD, Haugh KA, Pi R, Mothes W.
In vivo
Imaging-Driven Approaches to Study Virus
673
Dissemination and Pathogenesis. Annu Rev Virol. 2019;6(1):501-24. doi: 10.1146/annurev
-
674
virology
-101416-041429. PubMed PMID: 31283440; PubMed Central PMCID: PMC7217087.
675
9.
Pfeiffer JK. Innate host barriers to viral trafficking and population diversity: lessons learned
676
from poliovirus. Adv Virus Res. 2010;77:85-118. doi: 10.1016/B978-0-12-385034-8.00004-1.
677
PubMed PMID: 20951871; PubMed Central PMCID: PMC3234684.
678
10.
Fackler OT, Murooka TT, Imle A, Mempel TR. Adding new dimensions: towards an
679
integrative understanding of HIV
-1 spread. Nat Rev Microbiol. 2014;12(8):563-74. doi:
680
10.1038/nrmicro3309. PubMed PMID: 25029025; PubMed Central PMCID: PMC5687059.
681
11.
Sewald X, Ladinsky MS, Uchil PD, Beloor J, Pi R, Herrmann C, et al. Retroviruses use
682
CD169
-mediated trans
-infection of permissive lymphocytes to establish infection. Science.
683
2015;350(6260):563-7. doi: 10.1126/science.aab2749. PubMed PMID: 26429886; PubMed
684
Central PMCID: PMCPMC4651917.
685
12.
Uchil PD, Pi R, Haugh KA, Ladinsky MS, Ventura JD, Barrett BS, et al. A Protective Role
686
for the Lectin CD169/Siglec
-1 against a Pathogenic Murine Retrovirus. Cell Host Microbe.
687
2019;25(1):87-100 e10. doi: 10.1016/j.chom.2018.11.011. PubMed PMID: 30595553; PubMed
688
Central PMCID: PMC6331384.
689
13.
Kastenmuller W, Torabi
-Parizi P, Subramanian N, Lammermann T, Germain RN. A
690
spatially
-organized multicellular innate immune response in lymph nodes limits systemic
691
pathogen spread. Cell. 2012;150(6):1235-48. doi: 10.1016/j.cell.2012.07.021. PubMed PMID:
692
22980983; PubMed Central PMCID: PMCPMC3514884.
693
14.
Iannacone M, Moseman EA, Tonti E, Bosurgi L, Junt T, Henrickson SE, et al. Subcapsular
694
sinus macrophages prevent CNS invasion on peripheral infection with a neurotropic virus. Nature.
695
2010;465(7301):1079-83. doi: 10.1038/nature09118. PubMed PMID: 20577213; PubMed Central
696
P
MCID: PMCPMC2892812.
697
15.
Junt T, Moseman EA, Iannacone
M, Massberg S, Lang PA, Boes M, et al. Subcapsular
698
sinus macrophages in lymph nodes clear lymph-borne viruses and present them to antiviral B
699
cells. Nature. 2007;450(7166):110-4. doi: 10.1038/nature06287. PubMed PMID: 17934446.
700
16.
Honke N, Shaabani N, Cadeddu G, Sorg UR, Zhang DE, Trilling M, et al. Enforced viral
701
replication activates adaptive immunity and is essential for the control of a cytopathic virus. Nat
702
Immunol. 2012;13(1):51-7. doi: 10.1038/ni.2169. PubMed PMID: 22101728.
703
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprint
this version posted October 20, 2020.
;
https://doi.org/10.1101/2020.10.20.347427
doi:
bioRxiv preprint