VLDLR mediates alphavirus neuroinvasion through the blood-cerebrospinal fluid barrier

doi:10.21203/rs.3.rs-3404545/v1

Semliki Forest virus (SFV) is a neuropathogenic alphavirus which is of interest both as a model neurotropic alphavirus and as an oncolytic virus with proven potency in preclinical cancer models. The route of central nervous system (CNS) entrance of SFV is poorly understood but has been considered to occur through the blood-brain barrier. Here, we show that SFV primarily enters the CNS through the blood-cerebrospinal fluid barrier, and that VLDLR is crucial in enabling SFV infection of the choroid plexus epithelial cells.

Biological sciences/Microbiology/Virology/Alphaviruses

Health sciences/Pathogenesis/Infection

Biological sciences/Neuroscience/Blood–brain barrier

Alphavirus

SFV

blood-brain-barrier

blood-CSF barrier

neurovirulence

VLDLR

Semliki Forest virus (SFV) is an alphavirus, i.e., a group IV, positive-sense single-stranded RNA [+ ssRNA] virus in the Togaviridae family, which has been studied extensively as a model neurotropic alphavirus but also as an immunotherapeutic agent in preclinical tumor models including glioblastoma ^1–3. In vitro, SFV has the potency to infect and replicate in a wide range of cell types, originating from different hosts species. Recently, the Very Low-Density Lipoprotein Receptor (VLDLR) and Apolipoprotein E receptor 2 (ApoER2) were identified as entry receptors for Semliki Forest virus (SFV), Sindbis virus (SINV) and Easter equine encephalitis virus (EEEV) ⁴, and the structure of SFV in complex with VLDLR has been solved ⁵. In vivo, SFV shows very clear tropism towards the central nervous system (CNS) and in particular oligodendrocytes and neurons ⁶. Intravenous or intraperitoneal injections in laboratory mice lead to viral invasion of the central nervous system (CNS) and neuropathology of varying severity, depending on the SFV strain and age of the mice ^7–9. However, the cellular route of SFV neuroinvasion and the molecular mechanisms mediating CNS entry have not been described.

Here, we have employed Vldlr and Lrp8 (encoding ApoER2) KO mice to determine the mechanisms involved in SFV CNS entry and neuropathology. Contrary to the previously suggested model of CNS invasion through the blood-brain barrier (BBB), we demonstrate that SFV entry into the CNS occurs through the blood-cerebrospinal fluid (B-CSF) barrier at the choroid plexus (CP). Importantly, VLDLR is strictly required for the SFV infection of CP epithelial cells and thereby necessary for the subsequent neuroinvasion.

SFV strains A774 and SFV4 show distinctive VLDLR-dependency

To identify the SFV entry receptor, we used the human genome-wide Brunello CRISPR/Cas9 knockout library ¹⁰ in a screen on human osteosarcoma cells (HOS, ATCC, CRL-1543) infected with fully replicative SFV4 (Fig. 1A). While these cells are not a natural target for neurotropic SFV, they offer a highly SFV-permissive and easy to culture model needed for the screen. The janus kinase inhibitor ruxolitinib was used in combination with SFV4 to limit the effect of type-I interferon-mediated antiviral signaling thereby enhancing viral killing of infected cells and focusing the screening to crucial SFV4 entry factors. In line with the recent report by Clark et. al. ⁴, our screening identifies VLDLR as a receptor hit (Fig. 1B). A list of all significantly enriched hits with a false discovery rate (FDR) < 0.05 is presented in Fig. 1B. A full list of the results is provided in Data S1.

We next evaluated if VLDLR is a general receptor for SFV strains with different structural spike glycoproteins. For this, we selected SFV4 and the attenuated A774 strain (Fig. 1C), which differs in the structural spike glycoproteins E1 and E2 and (Fig S1A). In addition, we used the previously described chimeric A774-V4nstr strain ⁹, which has the structural glycoprotein genes from A774 and the replication-acting non-structural genes of SFV4 (Fig. 1C). This chimeric strain therefore allows identification of differences in infectivity between A774 and SFV4 that are specifically related to structural proteins, and therefore likely caused by differences in entry.

Consistent with SFV4 entry being VLDLR mediated, blocking VLDLR ligand-binding LDLR class A (LA) repeats with the monoclonal antibody (mAb) ¹¹ clone 1H5 in vitro protects HOS cells from SFV4-mediated lysis as compared to an isotype control antibody (Fig. 1D). Interestingly, VLDLR blocking gave no significant protection against A774 or A774-V4nstr (Fig. 1D). This indicates that, in contrast to the SFV4 strain, entry of SFV constructs carrying A774 structural proteins is not VLDLR-dependent. VLDLR blocking mAb clones 1H5 (binding LA repeats 2, 5 and 6 ¹¹) and 1H10 (binding repeats LA 3, 4, 5 and 6 ¹¹) inhibited SFV4 binding (Fig S1B and C). In contrast, mAb clone 5F3 (binding LA repeats 7, 8) gave no significant protection (Fig S1B and C). Our results are in line with recent results by Cao et. al., showing that VLDLR LA repeats 1, 2, 3 and 5 synergistically participate in SFV binding⁵. Infecting HOS cells with fluorescent reporter SFV4 (SFV4-dsRed) resulted in markedly lower replication (as detected by fluorescence intensity) in 1H10 and 1H5 pretreated cells (Fig. S1D). In line with A774 entry not being solely dependent on VLDLR, none of the mAbs used gave significant protection against A774 infection (Fig. S1C) and showed negligible reduction in fluorescent reporter A774-mCherry replication (Fig. S1E).

Further reinforcing the lower VLDLR-dependency of A774, the VLDLR-negative K562 cell line supported infection of both A774 and A774-V4nstr resulting in clearly increased virus titers in the culture medium (Fig. 1E). In contrast SFV4 titers in these cells were notably lower and below the originally added virus amount (Fig. 1E). Since VLDLR overexpression in K562 cells increased the titers of all SFV viruses (Fig. 1F), we conclude that SFV constructs carrying A774 structural proteins can indeed use VLDLR as a receptor but can readily use other receptor(s) in addition to VLDLR. On the other hand, entry of viruses with SFV4 structural proteins is dependent on VLDLR in vitro.

VLDLR is crucial for SFV neuroinvasion

To elucidate the role of VLDLR for SFV in vivo, we compared SFV pathology in wild-type (C57BL6/J, WT) and homozygous Vldlr knock-out mice (B6;129S7-Vldlr^tm1Her/J, Vldlr KO). Since the A774 replicase genes render the virus non-neurovirulent in adult mice⁶, we used the chimeric A774-V4nstr virus (instead of A774) in comparison with SFV4 to elucidate the effect of the spike glycoproteins in infection.

Both SFV4 and A774-V4nstr were neurovirulent when injected intravenously (IV) in WT mice as evidenced by poor survival due to neurological symptoms (Fig. 2A, Fig S2A) and presence of viral proteins throughout the brain parenchyma (Fig. 2B, Fig S2B) within 6 days after virus injection. The presence of virus and neurological symptoms was also accompanied by notable infiltration of CD45 + immune cells in the brain (Fig. 2D-F) indicating an SFV-induced severe immune-associated neuropathology. SFV4 was neurovirulent in the majority (9 of 12) of WT mice (Fig S2A), while A774-V4nstr was neuropathogenic in all (8 of 8) infected WT mice (Fig. 2A). This is in support of a wider receptor tropism mediated by A774 structural proteins. Both SFV4 and A774 entered the CNS in WT mice. In contrast, Vldlr KO mice were resistant to SFV-mediated encephalitis (Fig. 2A and Fig S2A) and showed low or no SFV protein staining in the brain at the end point on day 10 (Fig. 2C and Fig S2C). The drastic reduction of A774-V4nstr neuropathology in Vldlr KO mice was unexpected and indicates that VLDLR plays a non-redundant role during the onset of SFV neuropathology in mice. Direct intracranial administration of either SFV4 or A774-V4nstr led to rapid neuropathology in both WT and Vldlr KO mice (Fig S2D). This demonstrates that Vldlr KO mice are not generally protected against SFV, and that both virus strains have the potency to replicate once they enter the brains of Vldlr KO mice. VLDLR-expression correlated with SFV infectivity in certain brain cell populations. For example, Purkinje cells showed distinct VLDLR staining (Fig S2E) and were targets for SFV in WT mice but were protected against infection in Vldlr KO mice (Fig S2F). Interestingly, Vldlr KO mice were also susceptible for SFV when administered intranasally (Fig S2G) with evidence of the virus reaching not only the olfactory bulb but also deeper brain tissue including the medulla oblongata (Fig S2H).

The reduction of SFV neuropathology after systemic administration in Vldlr KO mice compared to WT mice led us to do a more detailed plaque titration analysis of viremia and infection in different parts of the brain (labelled as olfactory bulb, cortex, midbrain and cerebellum, Fig. 2G) as well as in the CSF 2, 3 and 4 days after IV injection of virus. The initial intravenous A774-V4nstr injection (1x10⁶ PFU) resulted in a second transient wave of viremia in WT mice as evidenced by virus in the blood samples on days 2 and 3 (Fig. 2H). Viremia was clearly reduced in Vldlr KO mice (Fig. 2H), which is likely evidence of a generally lower infectivity in these mice (including tissues outside the CNS). In WT mice, virus could be detected in all parts of the brain (in at least one of the analyzed mice) on days 3 and 4 (Fig. 2J-M). In Vldlr KO mice, virus was only detected in one of the brain samples (olfactory bulb, day 5, Fig. 2J). All the other Vldlr KO brain samples were negative, indicating that the virus did not enter into the CNS. This is in line with the lack of neuropathology observed in Vldlr KO mice. Interestingly, virus was detected in 2/5 CSF samples collected from WT mice on day 3 (Fig. 2I). This suggests that the virus has access to the CSF relatively early after intravenous virus injection. To increase the probability of detecting the virus from the very small volume CSF samples, we injected additional mice with higher dose of the virus (1x10⁷ PFU) and collected CSF. While this did not lead to higher proportion of samples being positive, we could detect higher titers in the positive samples (Fig S3A). This suggests a correlation between the amount of virus injected and the virus titer in the CSF. The CSF has been shown to naturally flow from the ventricular system ventrally to the basal cisterns and under the olfactory bulbs ¹²(Fig S3B). Matching this flow pattern, SFV protein was detected in the high VLDLR-expressing regions in the ventral side of the brain (near the olfactory bulb) at day 4 after IV A774-V4nstr injection (Fig S3C). This occurs before the onset of symptoms, which further supports that SFV first gains access to the ventricles and is subsequently carried to other regions in the brain.

Taken together, we conclude that VLDLR-expressing cells present a unique gateway for SFV neuroinvasion from the circulation, but that other receptors may contribute to infection within the CNS.

SFV does not efficiently penetrate the BBB

We next wanted to pinpoint the VLDLR-dependent entry route that SFV uses to access the CNS from the circulation. The first obvious potential entry route for SFV is through the blood brain-barrier (BBB), which has been previously suggested ¹³. However, according to data available from single-cell RNAseq database of gene expression in adult mouse brain ^14,15, mouse brain endothelial cells lack VLDLR expression (Fig S4A). In line with this, VLDLR staining was not observed in vascular structures in the brain of WT mice (Fig S4B). Interestingly, early signs of A774-V4nstr infection, by double-stranded (ds)RNA staining, was also noted at the BBB (as visualized by Type IV collagen surrounded structures) in both WT and Vldlr KO mice (Fig S4C-F). The presence of viral dsRNA was however not observed beyond the collagen IV layer neither in WT nor Vldlr KO mice, indicating a lack of entry into the brain parenchyma beyond the basement membrane. Both A774 and A774-V4nstr were capable of infecting isolated mouse brain endothelial cells, whereas SFV4 was not (Fig S4G). This result supports our results demonstrating A774 infection of VLDLR-negative K562 cells in vitro (Fig. 1E) and suggests that the A774 structural proteins mediate entry into VLDLR-negative brain endothelial cells.

ApoER2 has been suggested as an alternative SFV receptor and is expressed in brain endothelial cells (encoded by Lrp8 gene)^14,15 (Fig. S5A). However, homozygous Lrp8 knock-out mice (B6;129S6-Lrp8^tm1Her/J, Lrp8 KO) were only partly protected against intravenous A774-V4nstr, with 6/10 mice succumbing to infection (Fig. S5B) and the loss of Lrp8 did not prevent the virus from reaching the brain (Fig. S5C). This suggests that A774-V4nstr neuroinvasion from the circulation is less dependent on ApoER2 expressing cells than on VLDLR expressing cells. ApoER2-expressing brain endothelial cells can therefore not serve as major route of SFV neuroinvasion.

Taken together, our findings show that although endothelial cells of the BBB can be infected by SFV carrying the A774 glycoproteins (possibly via ApoER2), the infection does not spread beyond the basement membrane (Fig. S4C-F). As brain endothelial cells do not express VLDLR in WT mice (Fig. S4A), these cells cannot explain the lack of SFV-induced neuropathology in Vldlr KO mice. The BBB is therefore neither a highly effective nor a crucial entry route for SFV into the CNS.

VLDLR mediates SFV infection in the choroid plexus epithelial cells

Another possible route of infection from the circulation into the CNS is through the blood-CSF barrier. Unlike the tightly connected brain microvascular endothelial cells in the BBB, the fenestrated endothelial cell layer in the choroid plexus can serve as a viable bypass for SFV. Importantly, choroid plexus epithelial cells express VLDLR in WT mice (Fig. 3A-C) potentially supplying a pathway for SFV entry further into the CNS. Indeed, SFV proteins could be detected in the epithelial cells of the choroid plexus in WT mice early after intravenous A774-V4nstr injection (Fig. 3D). This finding, together with the presence of detectable levels of virus in the CSF of some mice (Fig. 2I and Fig S3) give strong evidence that the blood-CSF barrier is a more likely CNS entry pathway for SFV than the BBB. Lack of VLDLR in the epithelial cell layer of Vldlr KO mice could then explain the SFV resistant phenotype.

Staining of dsRNA corroborated A774-V4nstr replication in the WT choroid plexus (Fig. 3E). However, in Vldlr KO mice dsRNA could only be detected in the core of the choroid plexus (Fig. 3F), which includes endothelial cells and pericytes (Fig. 3G, H). This suggests that A774-V4nstr does not infect choroid plexus epithelial cells in Vldlr KO mice. Choroid plexus epithelial cells express low levels of ApoER2 (Lrp8) RNA (Fig. S6A). However, in line with previous observations in the chicken choroid plexus¹⁶, ApoER2 staining in WT epithelial cells shows an apical polarity (Fig. S6B-D). It is therefore likely that ApoER2 is not available as a receptor for virus invading from the basolateral side of the choroid plexus epithelial cell layer, thus explaining the crucial importance of VLDLR for SFV infection in these cells.

To summarize, our results pinpoint a crucial role of VLDLR in mediating SFV entry into choroid plexus epithelial cells, enabling spread into the CSF and further into the CNS. A schematic illustration of the route of SFV neuroinvasion based on our results is presented in Fig. 4.

The identity of the entry receptor for SFV has been a longstanding question in virology. Our data extend the results by Clark et. al. ⁴, validating VLDLR as the SFV entry receptor. More importantly, we demonstrate that VLDLR expression in the choroid plexus epithelial cells is crucial for SFV-mediated neuroinvasion and neuropathology. Thus, SFV binding to ApoER2 cannot compensate for a lack of VLDLR. Our results further point out an important difference in VLDLR utilization between the SFV strains A774 and SFV4. These SFV strains are not interchangeable and differences in their structural spike glycoproteins E1 and E2 results in notable difference in their VLDLR-dependency. SFV4 and A774 structural glycoproteins have been previously shown to have different affinity to bind heparan sulphate ¹⁷. In this context, it is not unexpected that they would also show different affinity to VLDLR or other possible entry receptors.

As reported previously, the low-density lipoprotein receptor ApoER2 has been identified as an alternative entry receptor for SFV ⁴. Unlike VLDLR, ApoER2 is preferentially expressed in the CNS ¹⁸, which is consistent with SFV infection within the brain parenchyma. Of note, ApoER2 was not enriched in our receptor screen. This can be explained by a strong VLDLR-dependency of the SFV4 strain which was used in the screen. Due to the nature of the screening (i.e., knocking out one gene per cell), the use of more promiscuous SFV constructs such as A774 would not be expected to lead to discovery of any significant receptor hits.

Vldlr KO mice have dramatically increased resistance to intravenously injected SFV4 and A774-V4nstr viruses (Fig. 2), even when (in the case of A774-V4nstr) the viral structural proteins allow the use of alternative entry receptors such as ApoER2. Interestingly, Lrp8 KO mice showed partial resistance to A774-V4nstr (Fig. S2I) which resembles SFV4 phenotype in WT mice (Fig. S2A). This suggests that the ability to use ApoER2 as a receptor increases SFV neuroinvasion but is not crucial as VLDLR is. Another potential alternative receptor candidate that we identified is OR10T2 (Olfactory Receptor Family 10 Subfamily T Member 2). While further validation of this alternative receptor hit is needed, it could explain the effectiveness of intranasal SFV delivery (Fig S2G).

The prevailing model for SFV infection prior to our report has been that SFV enters the CNS through the BBB ¹³. However, the lack of VLDLR expression in the brain endothelial cells ¹⁵ is likely to form a major barrier for VLDLR-dependent SFV clones such as SFV4. Other encephalitic alphaviruses have been reported to utilize transcytosis to pass through the endothelial cells in a receptor-independent manner¹⁹, but there is no evidence that SFV could utilize this route. Another feature of the BBB that may further limit entry of SFV is the structural barrier formed by astrocytes. Given that SFV has poor infectivity in astrocytes ⁶, it is reasonable to think that this is not a highly effective CNS entry route for SFV. Instead, CSF could serve a viable route into the brain parenchyma by facilitating access to ependymal cells and/or tanycytes as well as interneurons in perivascular space. Neural progenitor cells and neuroblasts in the subventricular zone could also serve as way for SFV to pass astrocytes layer. The possible role of these cells during SFV neuroinvasion requires further investigation.

The blood-CSF barrier as a viral entry into the CNS has been relatively poorly studied but has been reported for ZIKA virus ²⁰ and SARS-CoV-2 ²¹. Based on our results in Vldlr KO, Lrp8 KO and WT mice, we introduce a refined model of SFV neuroinvasion through the blood-CSF barrier. In our model, the passage through the endothelial cell layer can occur by ApoER2-mediated infection or by passive diffusion through fenestrations (Fig. 4). Passage through epithelial cells is strictly dependent on VLDLR, which (unlike ApoER2) in these cells is present also on the basolateral side (Fig. 4). In contrast to the BBB, the fenestrated endothelial layer and high VLDLR expression in the epithelial layer of the choroid plexus makes it an ideal entry route for SFV. Further studies are warranted to determine if SFV can also utilize circumventricular organs, which have permeable capillaries and direct access to neural tissues, as access points.

Taken together, our results show that VLDLR is crucial for SFV infectivity in choroid plexus epithelial cells and subsequent neuroinvasion through the blood-CSF barrier in mice. This present VLDLR with a defined and specific role in regulating CNS entry of SFV in vivo.

Data and code availability

This paper does not report original code. Any additional information required to reanalyze the data reported in this paper is available from the lead contact on request.

Ethical statement

The Swedish Work Environment Authority has approved the work with genetic modification of SFV (ID no. 202100-2932 v66a14 [laboratory] and v67a10 [mice]). All experiments regarding modified SFV were conducted under biosafety level 2. The local Animal Ethics Committee in Stockholm (13414/2020) approved the animal studies.

Author contribution

M.M., R.L., I.P., M.R., D.Y., A.D. and M.E. designed the experiments and/or helped with reagents and material. M.M., S.B., C.C, A.S, R.L., and I.P. conducted the experiments. M.M., A.D., and M.E. wrote the paper.

Declaration of interests

The Authors declare no competing interests.

Acknowledgements

We want to acknowledge Professor Gerald McInerney (Karolinska Institutet, Sweden), Andres Merits (University of Tartu, Estonia) and Ari Hinkkanen (University of Eastern Finland) for providing SFV-constructs. CRISPR Screen and the data analysis was done as a service by the CRISPR Functional Genomics unit at the SciLifeLab in Stockholm, Sweden.

Funding

This work was supported by the Knut and Alice Wallenberg Foundation [KAW 2019.0088]; the Swedish Cancer Society [190184Pj]; the Swedish Research Council [2019-01326]; the Swedish Childhood Cancer Foundation [PR2020-0167]. M.M. is a recipient of a Marie Curie fellowship from the EU (AVITAG, 707093).

Key resources table

REAGENT or RESOURCE	SOURCE	IDENTIFIER
Cell lines
BHK-21	ATCC	Cat# CCL-10
HOS	ATCC	Cat# CRL-1543
K562	ATCC	Cat# CCL-243
Primary mouse brain endothelial cells	this paper	N/A
HOS-Cas9/BFP	this paper	N/A
K5621-GFP/Fluc	this paper	N/A
K562-GFP/VLDLR	this paper	N/A
Viruses
pLV[Exp]-EGFP-EF1A>hVLDLR[NM_003383.5]	This paper	N/A
pCMV-SFV4	gift from Andres Merits	Reference 26
wtA7(74)	gift from Andres Merits	Reference 26
SFV4-d1EGFP	gift from Andres Merits	Reference 27
prA774-V4nstr	gift from Ari Hinkkanen	Reference 9
A774-mCherry	gift from Andres Merits	N/A
SFV4-dsRed	gift from Andres Merits	N/A
Primary antibodies
VLDLR antibody 1H5	GeneTex	Cat# GTX79551
VLDLR antibody 1H10	GeneTex	Cat# GTX79552
VLDLR antibody 5F3	GeneTex	Cat# GTX79550
Mouse IgG1 kappa Isotype Control	Thermo Fischer Scientific	Cat# 14-4714-82
rabbit anti-SFV	gift from Ari Hinkkanen	N/A
rat anti-CD45	BD Biosciences	Cat# 553076
goat anti-VLDLR	R&D Systems	Cat# AF2258
rabbit anti-collagen IV	Abcam	Cat# ab6586
mouse anti-dsRNA	Sigma-Aldrich	Cat# MABE1134
rabbit anti-AQP1	ThermoFischer Scientific	Cat# JM10-98
rabbit anti-Desmin	Abcam	Cat# ab15200
rabbit anti-PCP4	Novus	Cat# NBP1-80929
Secondary antibodies
donkey anti-rabbit-AF488	ThermoFischer Scientific	Cat# A-21206
donkey anti-rat-AF568	ThermoFischer Scientific	Cat# A-78946
donkey anti-goat-AF488	ThermoFischer Scientific	Cat# A-32814
donkey anti-rabbit-AF568	ThermoFischer Scientific	Cat# A-10042
donkey anti-mouse-AF647	ThermoFischer Scientific	Cat# A-31571
Reagents
RPMI-1640	Thermo Fischer Scientific	Cat# 21875-034
DMEM	Thermo Fisher Scientific	Cat# 41965-039
BHK-21 medium	ThermoFisher Scientific	Cat# 21710-025
OPTI-MEM	ThermoFisher Scientific	Cat# 31985-062
HEPES	Thermo Fisher Scientific	Cat# 15630056
Trptose Phosphate Broth	Teknova	Cat# T0800
Fetal Bovine serum (FBS)	ThermoFisher Scientific	Cat# 10500-064
L-glutamine	Thermo Fisher Scientific	Cat# 25030-081
Penicillin Streptomycin	Thermo Fisher Scientific	Cat# 15140-122
Sodium-pyruvate	Thermo Fisher Scientific	Cat# 11360-039
Cell Proliferation Kit I (MTT)	Roche	Cat#11465007001
OCT Mounting media	VWR Chemicals	Cat# 361603E
Fluoromount-G	Thermo Fisher Scientific	Cat# 00-4958-02
Hoechst 33342	Sigma-Aldrich	Cat# 14533
Bovine serum albumin (BSA)	Sigma-Aldrich	Cat# A9647
Phosphate-buffered saline (PBS)	Thermo Fisher Scientific	Cat# 10010023
Collagen-I	Sigma-Aldrich	Cat# C3867
Heparin	Sigma-Aldrich	Cat# H3149-KU50
ECGS	Sigma-Aldrich	Cat# E2759
Ruxolitinib phosphate	Selleck Chemicals	Cat# S5243
Cell proliferation kit I	Merck	Cat# 11465007001
collagenase/dispase	Merc	Cat# 10269638001
DNase	Sigma-Aldrich	Cat# D4513
Puromycin Dihydrochloride	Thermo Fisher Scientific	Cat# A1113803

Experimental models
WT mice	Charles River Laboratories	C57BL/6J
Vldlr KO	The Jakcson Laboratory	B6;129S7-Vldlrtm1Her/J
Lrp8 KO	The Jakcson Laboratory	B6;129S6-Lrp8tm1Her/J,
Software and algorithms
GraphPad Prism Version 7 or higher	Dotmatics	https://www.graphpad.com/
Fiji/ImageJ 1.21	NIH	https://fiji.sc/
Adobe Illustrator 2023	Adobe inc.	https://www.adobe.com/se/products/illustrator
ZEN Microscopy Software	ZEISS	https://www.zeiss.com/microscopy/en/products/software/zeiss-zen.html
Leica Application Suite X	Leica Microsystems	https://www.leica-microsystems.com/products/microscope-software/p/leica-las-x-ls/
MARS Data Analysis Software	BMG labtech	https://www.bmglabtech.com/en/microplate-reader-software/

Resource availability

Lead contact

Further information and requests for resources and reagents should be directed to and will be fulfilled by the lead contact, Miika Martikainen ([email protected])

Materials availability

Materials generated in this study will be made available upon request, but we may require a completed Materials Transfer Agreement if there is potential for commercial application.

Method details

Cell line and viruses

HOS (ATCC, CRL-1543) and K562 (ATCC, #CCL-243) cells were cultured in Gibco RPMI-1640 (Thermo Fischer Scientific, Waltham, MA, #21875-034) supplemented with 10% Gibco heat-inactivated FBS (Thermo Fischer Scientific, #10500-064), 10 U/mL penicillin-streptomycin (Thermo Fisher Scientific, #15140-122), and 1 mM sodium pyruvate (Thermo Fisher Scientific, #11360-039). To create hVLDLR-expressing K562 cell line, K562 were transduced by lentivirus expressing GFP and VLDLR (pLV[Exp]-EGFP-EF1A>hVLDLR[NM_003383.5], VectorBuilder). Control cells were transduced with lentivirus expressing GFP and Firefly luciferase ²³. After transduction the cells were sorted based on GFP expression using BD FACSMelody cell sorter.

Fully replicative viruses SFV4 (pCMV-SFV4) ²⁴, wtA7(74)²⁴ and SFV4-d1EGFP ²⁵ , A774-mCherry and SFV4-dsRed were obtained from Andres Merits (University of Tartu, Estonia). Chimeric prA774-V4nstr ⁹ was gift from Ari Hinkkanen (University of Eastern Finland). Viruses were produced and titrated in BHK-21 cells as described previously in ². Full length sequences of all viruses are available upon request.

CRISPR/Cas9 library screen

Generation of HOS-Cas9/BFP and transduction with lentiviral human genome-wide Brunello CRISPR-library¹⁰ as service by SciLifeLab CRISPR Functional Genomics unit (Karolinska Intitute, Stockholm, Sweden). 40 000 000 cells/sample were used in the screen to maintain full coverage of the library. For the screening cells were given either ruxolitinib alone (10µM, Selleck Chemicals, Planegg, Germany, #S5243) or ruxolitinib and SFV4-d1-EGFP (MOI=50). Two days later the surviving cells were harvested sgRNA expression was analyzed by MAGeCK analysis ²⁶ by SciLifeLab.

VLDLR blocking experiments

HOS cells were seeded on 96 well-plate (10 000 cells/well). On the next day medium was replaced with 50µl VLDLR blocking mAbs 1H5, 1H10, 5F3 (GeneTex Irvine, CA, #GTX79551, #GTX79552 and #GTX79550) or IgG1 isotype control antibody (Thermo Fischer Scientific, #14-4714-82) diluted in complete RPMI-1640 medium at 100µg/ml. 2h later, 50µl of virus (MOI=0.1) diluted in complete RPMI-1640 medium was added and cells. 48h later, cell viability was measured with Roche Cell Proliferation Kit I MTT assay (Merck, Darmstadt, Germany, #11465007001) according to manufacturer’s instructions.

Alternatively, HOS cells treated with 1H5, 1H10, 5F3 antibodies (as above) were infected with red fluorophore expressing SFV viruses A774-mCherry or SFV4-dsRed (kind gift form prof. Andres Merits, Tartu, Estonia) using MOI 0.1 and the fluorescence intensity was measured with CLARIOstar plate reader and analyzed with MARS software (BMG Labtech, Ortenberg, Germany)

SFV replication kinetics in K562 cell line

200 000 K5621-GFP/Fluc or K562-GFP/VLDLR were seeded on 12 well-plate and infected with SFV4, A774 or A774-V4nstr virus (MOI=0.01) at 1ml complete RPMI-164 medium. 200µl of medium sample was collected at indicated time points and virus amount was quantified by plaque titration in BHK-21 cells.

Infection of mice

For analysis of neurovirulence, adult (>6-week-old) female WT (C57BL/6J, Charles River Laboratories, Wilmington, MA), VLDLR^-/- (B6;129S7-Vldlr^tm1Her/J, The Jakcson Laboratory, Bar Harbor, ME) or ApoER2^-/- (B6;129S6-Lrp8^tm1Her/J, The Jakcson Laboratory) were used. For intravenous (IV) injection, 1x10⁶plaque-forming units (PFUs, as titrated in BHK-21 cells⁵) of virus was injected into the tail vein in total volume of 100µl of phosphate-buffered saline (PBS). Intranasal injections of virus (1x10⁶PFUs in 10µl PBS) were done with a pipette into the left nostril under isoflurane anesthesia. Intracranial injections of virus (1000 PFU in 2µl PBS) were done 1 mm anterior and 1.5 mm right from bregma at 3-mm depth using a Hamilton syringe and stereotactic equipment (AgnTho’s, Sweden). Mice were monitored daily and sacrificed either upon onset of neurological symptoms (ataxia, paralysis, hunched posture or over 20% loss of body weight) or 10 days after virus injection.

Analysis of SFV replication in different brain regions, blood, and CSF

Adult female WT (C57BL/6J, Charles River Laboratories) or VLDLR^-/- (B6;129S7-Vldlrtm1Her/J, The Jakcson Laboratory) were infected intravenously with A774-V4nstr (1x10⁶ PFU in PBS) followed by collection of samples on days 2, 3 and 4 after virus injection. Mice were put under terminal anaesthesia and CSF was collected with small glass capillary via cisterna magna followed by blood collection via cardiac puncture. Immediately after this the mice were perfused with 10ml PBS through left ventricle and brains were collected. Brains were dissected to smaller pieces corresponding to specific regions (according to schematics in Fig.2G) olfactory bulb, cerebellum, cortex (including hippocampus, thalamus, and hypothalamus) and midbrain (including pons and medulla regions). Samples were snap-frozen and stored in -20°C.

Blood and CSF were directly diluted into BHK-21 medium (Thermo Fisher Scientific, #21710-025) and used for plaque titration. Frozen brain pieces were suspended into 200µl of OPTI-MEM (Thermo Fisher Scientific, #31985-062) and homogenized with a disposable tissue grinder (Fisher Scientific, # 13236679) inside 1.5ml tube. The resulting homogenate was centrifuged max speed 5min at RT, and the supernatant was used for plaque titration.

Immunostainings

Brains from sacrificed mice were snap-frozen in 2-methylbutane on dry ice, embedded to OCT embedding matrix (VWR Chemicals, Radnor, PA, Cat# 361603E) and cut into 7 μm sections with a cryostat and mounted on Superfrost Plus microscope slides (Thermo Scientific, #J1800AMNZ).

Sections were fixed either with methanol or acetone in -20°C, washed with PBS and then blocked in PBS containing 3% BSA (1h RT). Primary antibody incubation was done overnight at 4°C (with antibodies listed in key resources table). Sections were washed with PBS and incubated with fluorophore-conjugated secondary antibodies (listed in key resource table) for 2 hours at room temperature. Sections were the washed in PBS, counterstained with Hoechst 33342 (Sigma-Aldrich, #14533), and mounted using Fluoromount-G (Thermo Fisher Scientific, #00-4958-02). Fluorescent images were captured with a Zeiss Axioimager microscope (Zeiss, Oberkochen, Germany) or Leica SP8 confocal microscope (Leica Microsystems, Wetzlar, Germany).

Isolation and infection of mouse brain endothelial cells (BECs)

Brains of 10 adult female C57BL6/J mice (Charles River) were harvested. Meninges were removed by rolling the brains on sterile Whatman paper, after which brains were minced with scalpel blades, pooled together, and suspended in 50ml of Buffer A (150mM NaCl, 5mM KCl, 2mM CaCl₂, 2.6mM MgCl₂, 15mM Hepes, 1% BSA in water).

Suspension was centrifuged (5min, 1200 rpm, RT) and the pellet was suspended in equal volume (compared to the volume of the pellet) of 0.75% Collagenase (Type 2, LS004176, Worthington) in Buffer A. Suspension was incubated 1h at 37°C after which the digestion was stopped by adding buffer A up to 50ml.

After centrifugation (10 minutes at 1200 rpm at 4°C) the pellet was suspended in 30ml of 25% BSA in PBS followed by another round of centrifugation (20 minutes at 2600 rpm at 4°C). The supernatant, including a layer of myelin on top, was removed and the pellet suspended in 4 ml of Buffer A. 40µl of 1% collagenase/dispase (Merc, #10269638001, in buffer A) was added and suspension was incubated 15 minutes at 37°C. After this, 4µl of DNase (Sigma-Aldrich, # D4513, 1 mg/ml Buffer A) was added and the suspension was incubated additional 2 minutes at 37°C, gently shaking to resolve any clumps.

Digestion was stopped by adding Buffer A up to 50 ml and centrifuged for 10 minutes at 1200 rpm at 4°C. The resulting cell pellet was suspended in EC medium [DMEM (Thermo Fisher Scientific, #41965-039) complemented with 10% FBS (ThermoFisher Scientific, #10500-064), 50µg/ml ECGS (Sigma-Aldrich, #E2759), 100µg/ml Heparin (Sigma-Aldrich, #H3149-KU50) and Pen/Strep (Thermo Fisher Scientific, #15140-122)] and plated on collagen-coated (Sigma-Aldrich, #C3867) 6well plate (split evenly to all wells) and let grow in 37°C incubator (5% CO₂). On the following day, cells were washed with PBS and new EC medium supplemented with 4ug/ml puromycin was added.

After 2 days of puromycin (Thermo Fisher Scientific, #A1113803) selection (4µg/ml) cells were infected by adding SFV4, A774 or A774-V4nstr virus at MOI=0.01 (diluted to EC medium without puromycin). Medium samples were collected 24 hours after infection for virus titration on BHK-21 cells.

Statistical analysis

All statistical analysis was performed using GraphPad Prism software.

Martikainen, M. et al. MicroRNA-Attenuated Clone of Virulent Semliki Forest Virus Overcomes Antiviral Type I Interferon in Resistant Mouse CT-2A Glioma. Journal of virology 89, 10637–10647, doi:10.1128/jvi.01868-15 (2015).
Martikainen, M. et al. IFN-I-tolerant oncolytic Semliki Forest virus in combination with anti-PD1 enhances T cell response against mouse glioma. Molecular therapy oncolytics 21, 37–46, doi:10.1016/j.omto.2021.03.008 (2021).
Ramachandran, M. et al. Safe and Effective Treatment of Experimental Neuroblastoma and Glioblastoma Using Systemically Delivered Triple MicroRNA-Detargeted Oncolytic Semliki Forest Virus. Clinical cancer research: an official journal of the American Association for Cancer Research 23, 1519–1530, doi:10.1158/1078-0432.ccr-16-0925 (2017).
Clark, L. E. et al. VLDLR and ApoER2 are receptors for multiple alphaviruses. Nature 602, 475–480, doi:10.1038/s41586-021-04326-0 (2022).
Cao, D., Ma, B., Cao, Z., Zhang, X. & Xiang, Y. Structure of Semliki Forest virus in complex with its receptor VLDLR. Cell 186, 2208–2218.e2215, doi:10.1016/j.cell.2023.03.032 (2023).
Fragkoudis, R. et al. Neurons and oligodendrocytes in the mouse brain differ in their ability to replicate Semliki Forest virus. Journal of neurovirology 15, 57–70, doi:10.1080/13550280802482583 (2009).
Fazakerley, J. K., Pathak, S., Scallan, M., Amor, S. & Dyson, H. Replication of the A7(74) strain of Semliki Forest virus is restricted in neurons. Virology 195, 627–637, doi:10.1006/viro.1993.1414 (1993).
Oliver, K. R., Scallan, M. F., Dyson, H. & Fazakerley, J. K. Susceptibility to a neurotropic virus and its changing distribution in the developing brain is a function of CNS maturity. Journal of neurovirology 3, 38–48, doi:10.3109/13550289709015791 (1997).
Tuittila, M. T., Santagati, M. G., Röyttä, M., Määttä, J. A. & Hinkkanen, A. E. Replicase complex genes of Semliki Forest virus confer lethal neurovirulence. Journal of virology 74, 4579–4589, doi:10.1128/jvi.74.10.4579-4589.2000 (2000).
Doench, J. G. et al. Optimized sgRNA design to maximize activity and minimize off-target effects of CRISPR-Cas9. Nature biotechnology 34, 184–191, doi:10.1038/nbt.3437 (2016).
Ruiz, J. et al. The apoE isoform binding properties of the VLDL receptor reveal marked differences from LRP and the LDL receptor. Journal of lipid research 46, 1721–1731, doi:10.1194/jlr.M500114-JLR200 (2005).
Decker, Y. et al. Magnetic resonance imaging of cerebrospinal fluid outflow after low-rate lateral ventricle infusion in mice. JCI insight 7, doi:10.1172/jci.insight.150881 (2022).
Fazakerley, J. K. Semliki forest virus infection of laboratory mice: a model to study the pathogenesis of viral encephalitis. Archives of virology. Supplementum, 179–190, doi:10.1007/978-3-7091-0572-6_16 (2004).
He, L. et al. Single-cell RNA sequencing of mouse brain and lung vascular and vessel-associated cell types. Scientific data 5, 180160, doi:10.1038/sdata.2018.160 (2018).
Vanlandewijck, M. et al. A molecular atlas of cell types and zonation in the brain vasculature. Nature 554, 475–480, doi:10.1038/nature25739 (2018).
Stockinger, W. et al. The low density lipoprotein receptor gene family. Differential expression of two alpha2-macroglobulin receptors in the brain. The Journal of biological chemistry 273, 32213–32221, doi:10.1074/jbc.273.48.32213 (1998).
Ferguson, M. C. et al. Ability of the Encephalitic Arbovirus Semliki Forest Virus To Cross the Blood-Brain Barrier Is Determined by the Charge of the E2 Glycoprotein. Journal of virology 89, 7536–7549, doi:10.1128/jvi.03645-14 (2015).
Reddy, S. S., Connor, T. E., Weeber, E. J. & Rebeck, W. Similarities and differences in structure, expression, and functions of VLDLR and ApoER2. Molecular neurodegeneration 6, 30, doi:10.1186/1750-1326-6-30 (2011).
Salimi, H. et al. Encephalitic Alphaviruses Exploit Caveola-Mediated Transcytosis at the Blood-Brain Barrier for Central Nervous System Entry. mBio 11, doi:10.1128/mBio.02731-19 (2020).
Kim, J. et al. Zika virus infects pericytes in the choroid plexus and enters the central nervous system through the blood-cerebrospinal fluid barrier. PLoS pathogens 16, e1008204, doi:10.1371/journal.ppat.1008204 (2020).
Pellegrini, L. et al. SARS-CoV-2 Infects the Brain Choroid Plexus and Disrupts the Blood-CSF Barrier in Human Brain Organoids. Cell stem cell 27, 951–961.e955, doi:10.1016/j.stem.2020.10.001 (2020).
Ximerakis, M. et al. Single-cell transcriptomic profiling of the aging mouse brain. Nature neuroscience 22, 1696–1708, doi:10.1038/s41593-019-0491-3 (2019).
Zhang, L. et al. Pleiotrophin promotes vascular abnormalization in gliomas and correlates with poor survival in patients with astrocytomas. Science signaling 8, ra125, doi:10.1126/scisignal.aaa1690 (2015).
Ulper, L., Sarand, I., Rausalu, K. & Merits, A. Construction, properties, and potential application of infectious plasmids containing Semliki Forest virus full-length cDNA with an inserted intron. Journal of virological methods 148, 265–270, doi:10.1016/j.jviromet.2007.10.007 (2008).
Rausalu, K. et al. Properties and use of novel replication-competent vectors based on Semliki Forest virus. Virology journal 6, 33, doi:10.1186/1743-422x-6-33 (2009).
Li, W. et al. MAGeCK enables robust identification of essential genes from genome-scale CRISPR/Cas9 knockout screens. Genome biology 15, 554, doi:10.1186/s13059-014-0554-4 (2014).

There is NO Competing Interest.

SFVRuxvsControlRux.genesummary.xlsx
Dataset 1
FigureS1S6.docx

VLDLR mediates alphavirus neuroinvasion through the blood-cerebrospinal fluid barrier

Status:

Version 1

Abstract

Figures

Introduction

Results

SFV strains A774 and SFV4 show distinctive VLDLR-dependency

VLDLR is crucial for SFV neuroinvasion

SFV does not efficiently penetrate the BBB

VLDLR mediates SFV infection in the choroid plexus epithelial cells

Discussion

Declarations

STAR Methods

References

Additional Declarations

Supplementary Files

Status:

Version 1