advances.sciencemag.org/cgi/content/full/5/10/eaaw8651/DC1

Supplementary Materials for

HIV-1 Gag specifically restricts PI(4,5)P2 and cholesterol mobility in living cells

creating a nanodomain platform for virus assembly

C. Favard, J. Chojnacki, P. Merida, N. Yandrapalli, J. Mak, C. Eggeling*, D. Muriaux*

*Corresponding author. Email: delphine.muriaux@irim.cnrs.fr (D.M.); christian.eggeling@rdm.ox.ac.uk (C.E.)

Published 2 October 2019, Sci. Adv. 5, eaaw8651 (2019) DOI: 10.1126/sciadv.aaw8651

The PDF file includes:

Sections S1 and S2. Characterization of fluorescent lipid analogs Section S3. MHC-I mobility at HIV-1 assembly sites in NL4.3 Gag-iGFP HIV-1–infected Jurkat T cells Section S4. STED microscope calibration Section S5. PI(4,5)P2 clustering and trapping by full-length HIV-1 Gag on biomimetic membranes Section S6. Budding efficiency comparison in Jurkat T cells transfected with Gag.eGFP or with a mixture of Gag.eGFP/Gag (ratio, 1:3) Section S7. Cumulative frequency distributions observed in infected cells for the different lipids Section S8. Cumulative frequency distributions observed in transfected cells for the different lipids Fig. S1. Fluorescent lipid analog structures. Fig. S2. Fluorescent lipid analog mobility in Jurkat T cells. Fig. S3. MHC-I mobility at HIV-1 assembly sites in NL4.3 Gag-iGFP HIV-1–infected Jurkat T cells. Fig. S4. STED microscope calibration results. Fig. S5. Changes in the lateral mobility of ATTO647N-PI(4,5)P2 upon addition of Gag on SLBs. Fig. S6. Jurkat T cell coelectroporation with Gag.eGFP and Gag plasmids. Fig. S7. Cumulative frequency distributions of diffusion coefficient observed in HIV-1–infected T cells. Fig. S8. Cumulative frequency distributions of diffusion coefficient observed in HIV-1 Gag-transfected T cells. Legend for movie S1 References (62, 63)

Other Supplementary Material for this manuscript includes the following: (available at advances.sciencemag.org/cgi/content/full/5/10/eaaw8651/DC1)

Movie S1 (.avi format). Drift stabilized time lapse movie of a representative NL4.3 Gag-iGFP HIV-1–infected Jurkat T-cell showing already present and newly developing virus assembly sites.

Supplementary Materials Sections S1 and S2. Characterization of fluorescent lipid analogs Scanning STED FCS requires dedicated lipid analogs with photostable and bright dye labels

such as Atto647N or KK114. Figure S1 shows the structures of the lipid analogs used throughout the present study. These analogs have been used multiple times before and in their

cases the dye label shown to have minimal influence on the lipids’ membrane interaction

dynamics (34–37) as confirmed by using classical spot FCS (fig. S2), highlighting that the

mobility of these lipid analogs in CD4+ T-cells was comparable between different dye labels,

specifically to the mobility observed with corresponding BODIPY labelled lipids that we

previously used to monitor the effect of Gag self-assembly on model membranes (10)

Fig. S1. Fluorescent lipid analog structures. From left to right, structures of all the fluorescent sterols and lipid analogs used in this study, a- KK114 Chol, b- TopFluor Cholesterol, c- Bodipy TMR C16 PI(4,5)P2, d- Atto647N- PI(4,5)P2, e- TopFluor PI(4,5)P2, f- TopFluor SM, g- Atto647N SM, h-Atto647N-PE, i- KK114-PE.

Fig. S2. Fluorescent lipid analog mobility in Jurkat T cells. a – Transmission (upper part) and fluorescent (lower part) image of Jurkat T-cells labelled with Atto647N-SM or KK114-

Chol (scalebar=5 µm). b- Examples of normalized correlograms and their corresponding fit using equation 1 obtained in the labelled Jurkat T-cells for KK114-Chol (red line) and

Atto647N-SM (black line). c- Box-plot of the apparent diffusion coefficients observed in Jurkat T-cells labelled with different lipid analogs for PI(4,5)P2 (red boxes), Cholesterol

(green boxes) and sphingomyelin (blue boxes) and measured using confocal spot-FCS (empty

boxes) and compared to scanning confocal FCS (diagonal pattern filled boxes).

Section S3. MHC-I mobility at HIV-1 assembly sites in NL4.3 Gag-iGFP HIV-1–infected Jurkat T cells In order to further reinforce the idea that our experiments are not observing fully

assembled/budded viruses on the cell membrane, we have performed an experiment

measuring MHC-I mobility at HIV-1 virus assembly sites in infected Jurkat T-cells. For the

observation of MHC-I mobility, NL4.3 Gag-iGFP infected cells were stained for MHC-I in

suspension at 16 °C using W6/32 anti-MHC-I Fab fragments and anti-human Abberior STAR

RED (KK114) conjugated Fab fragments for 1 h each in 0.5% BSA/L-15 Medium (3).

Mobility analysis inside and outside selected virus assembly sites shows that there is no

significant MHC-I mobility change at these sites compared to areas outside and uninfected

control cells. In a previous study (3), we shown that MHC-I mobility is highly reduced on

fully assembled and budded viruses (Dmedian = 0.0027 μm2/s). These results indicate that virus

assembly sites analysed in this study do not represent fully assembled and budded virus

particles. Small mobility reduction seen between inside and outside assembly site areas may

be due to the early stages of the virus assembly beginning to affect MHC-I diffusion in these

areas.

Fig. S3. MHC-I mobility at HIV-1 assembly sites in NL4.3 Gag-iGFP HIV-1–infected Jurkat T cells. Median MHC-I diffusion coefficient (D) was determined by sSTED-FCS measurements of 20 sites each from two independent virus preparations and infections. Box

and whisker plots (horizontal line – median, box – 25-75 % percentiles or interquartile range

(IQR) and whiskers – 10-90 % measurements) shows D values inside assembly sites (red),

outside assembly sites (green) and in non-infected control cells (blue). Statistical significance

was assessed by Wilcoxon rank-sum test.

Section S4. STED microscope calibration

The diameters of the observation spots employed in the scanning STED-FCS was estimated at

37 °C following a standard calibration procedure on supported lipid bilayers (SLBs) (61),

comparing transit times of freely diffusing fluorescent lipid (KK114-DPPE) in confocal and

STED microscopy modes (fig. S4).

Fig. S4. STED microscope calibration results. (a) Mean ± SD of observation spot diameters at different STED laser powers was determined by STED-FCS calibration measurements of

KK114-DPPE diffusion in supported lipid bilayers (SLBs). Results represent 10

measurements each from 2 separate preparations. For the virus assembly sites measurements,

a minimum STED laser power (90 mW) was set to generate an observation spot diameter (100

nm) below the diameter of HIV-1 particles (≈ 140 nm diameter) and virus assembly sites,

while maintaining as high as possible signal-to-noise ratio and minimising the laser light

exposure of live Jurkat T cells.

Section S5. PI(4,5)P2 clustering and trapping by full-length HIV-1 Gag on biomimetic membranes

Yandrapalli et al.(10) previously reported the ability of purified full length Gag to segregate

PI(4,5)P2 and cholesterol while self-assembling on supported lipid bilayers (SLB). Here, we

checked the ability of Gag to induce PI(4,5)P2 clustering and trapping on SLBs with the basic

lipid composition described in Yandrapalli et al.(10) by imaging the change in ATTO647N-

PI(4,5)P2 fluorescence before and after addition of Gag on SLBs (fig. S2a and movie S1). As

previously reported, we observed the generation of ATTO647N-PI(4,5)P2 enriched domains

few minutes after injection of recombinant full length Gag (fig. S5a). Concomitantly, confocal

line scanning FCS (sFCS) was performed on SLBs in order to monitor the changes in lateral

PI(4,5)P2 mobility induced by Gag self-assembly. We first measured the lateral diffusion of

ATTO647N-PI(4,5)P2 before Gag injection (fig. S5b, blue box) and compared it to the lateral

diffusion observed after Gag injection inside (fig. S5b, red box) and outside (fig. S5b, green

box) the generated Gag aggregation sites. We observed a significant decrease in ATTO647N-

PI(4,5)P2 lateral mobility at (Da = 0.15 µm2.s-1) compared to outside (Da = 1.1 µm2.s-1) the

sites or before the addition of Gag (Da = 0.95 µm2.s-1), highlighting strong trapping of

ATTO647N-PI(4,5)P2 in lipid domains generated by Gag.

We then applied the confinement index calculation to determine the enrichment of

ATTO647N-PI(4,5)P2 in Gag induced clusters. We obtained values in line with either HIV-1

infected cells or HIV-1 Gag.eGFP transfected cells. (Figs. 2 and 3)

Fig. S5. Changes in the lateral mobility of ATTO647N-PI(4,5)P2 upon addition of Gag on SLBs. (a) Confocal images of PC/PS/PI(4,5)P2 supported lipid bilayers labelled with ATTO647N-PI(4,5)P2 before (upper part) and 3 minutes after addition of the full length HIV-

1 Gag protein. (b) Box plot of the diffusion coefficients of ATTO647N-PI(4,5)P2 observed before addition of Gag (blue) and 3 minutes after addition of Gag inside (red) or outside

ATTO647N-PI(4,5)P2 clusters (green), (boxes are median, first and third quartile and 10-90%

whiskers). Values of the confinement indices of PI(4,5)P2 analog in SLB for the three

different diffusion regimes (slow Da < 0.1 µm².s-1, intermediates 0.1 < Da < 1µm².s-1) and fast

(Da > 1 µm².s-1). As in the case of infected and transfected cells, ATTO647N-PI(4,5)P2 is

strongly enriched in the clusters observed 3 min after addition of Gag.

Section S6. Budding efficiency comparison in Jurkat T cells transfected with Gag.eGFP or with a mixture of Gag.eGFP/Gag (ratio, 1:3) We attempted to transfect Jurkat T cells, using electroporation, with a mixture of tagged and

untagged Gag expressing plasmids, similarly to studies done on highly transfectable adherent

cell lines, such as HeLa or 293T cells (see Jouvenet et al., (62) and Ivanchenko et al., (32)),

but never on CD4 T cells. When electroporated with 2 plasmids we observed, using anti-CA

(p24) immunoblots, that one of the plasmids is better expressed than the other: here, it is the

pGag plasmid – producing unlabelled Gag proteins (fig. S6a). Moreover, with a ratio of 1 to 3

Gag.eGFP/Gag expressing plasmids we were not able to detect GFP(+) cells producing

fluorescent VLPs and none of them were producing detectable virus assembly sites. This lack

of production of fluorescent dots (virus assembly) at their cell surface thus prevents us from

doing STED-FCS in and out of the virus assembly sites. In these conditions Gag.eGFP stays

diffuse in the cytosol even 40h post-transfection, indicating probably that the unlabelled Gag

protein is assembling preferentially, leaving the Gag.eGFP poorly incorporated into the viral

budding site (fig. S6b). This data indicates that we are unable to do a plasmid mixture

experiments on electroporated Jurkat T cells.

Nonetheless, we believe that Gag.eGFP only experiment is still useful in the context of our

study. This is because, while constructs such as pKHIV.eGFP or NL43.eGFP indeed require a

rescue (50% of untagged Gag) to get a fully infectious virus, early studies using these

constructs suggest that the infectivity defects in Gag.eGFP-only particles arise from incorrect

Gag processing (maturation) following assembly rather than Gag assembly itself (63).

Furthermore, in a recently study (33), we have shown that in Jurkat T cells expressing 100%

mEOS2-tagged(i)Gag, we were able to monitor at the single molecule level, the kinetic of

Gag assembly and it was found to resemble the one described in HeLa cells (5 min of

assembly + 10 min of particle residence at the cell membrane and release).

We would also like to emphasise that similar lipid mobility results were also obtained using

fully infectious NL43 Gag.iGFP plasmid. This plasmid is described in the study of Hubner et

al. (31), which characterize fully infectious NL4.3 Gag.iGFP virus particles. Although the

authors do not comment on the assembly and release competency of the particles they indicate

that the construct and expression of the proviral clone HIV Gag-iGFP, with the GFP inserted

between MA and CA and with 2 protease cleavage sites have normal expression, viral

production and infectivity, very close to wild-type HIV-1. The authors show that NL4.3

Gag.iGFP retains full infectivity as wild type and that the distribution of Gag on cells surface

is the same (by confocal microscopy). They also indicate correct assembly/budding pathway.

This make us confident that even if maturation and infectivity defects exist when using 100%

GFP-tagged Gag, this tagged protein is still useful for our measurements which target only an

early assembly phase.

Fig. S6. Jurkat T cell coelectroporation with Gag.eGFP and Gag plasmids. (a) Jurkat T cells were electroporated with empty plasmid (mock), GFP, Gag.eGFP or a mixture of

Gag.eGFP+Gag (ratio 1:3) expressing plasmids and analysed 40 hours post-electroporation,

as seen in images using low magnification fluorescence microscopy. (b) A Western blot anti-CAp24 of the electroporated Jurkat T cells in the different conditions. It can be seen that in

the mixture condition, Gag is highly expressed as compared to Gag.eGFP, lowering the

probability to get a T cell producing GFP(+) viruses. (c) Confocal microscopy images of electroporated Jurkat T cells. Cells were rinsed, resuspended in L15 transparent media,

adhered on Poly-L coverslip at 37°C for 1h and imaged by confocal microscopy. A typical

living GFP(+) T cell is shown for the condition Gag.eGFP and for the condition

Gag.eGFP+Gag (ratio 1:3) as indicated. Results demonstrate that, viral assembly sites

produced by the T cell can be observed only in the condition with Gag.eGFP only. It is never

the case for Jurkat Tcells electroporated with the mixture Gag.eGFP/Gag. LUT scale:

Blue/purple: low fluorescence intensity, yellow: high fluorescence intensity.

Section S7. Cumulative frequency distributions observed in infected cells for the different lipids

Fig. S7. Cumulative frequency distributions of diffusion coefficient observed in HIV-1–infected T cells. Cumulative frequency distributions of values of apparent diffusion coefficients Da measured inside (red) and outside (blue) assembly sites of infected cells, and

in non-infected cells (green) for the different fluorescent lipid analogs, ATTO647N-PI(4,5)P2

(a), Cholesterol (Chol-PEG-KK114) (b), sphingomyelin (ATTO647N-SM) (c) and DPPE analogs (d).

Section S8. Cumulative frequency distributions observed in transfected cells for the different lipids

Fig. S8. Cumulative frequency distributions of diffusion coefficient observed in HIV-1 Gag-transfected T cells. Cumulative frequency distributions of the Da values measured inside (red) and outside (blue) assembly sites of transfected cells, and in non-transfected cells

(green) for the different lipids, ATTO647N-PI(4,5)P2 (a), Cholesterol (Chol-PEG-KK114) (b), sphingomyelin (ATTO647N-SM) (c) and DPPE analogs (d).

Movie S1. Drift stabilized time lapse movie of a representative NL4.3 Gag-iGFP HIV-1–infected Jurkat T-cell showing already present and newly developing virus assembly sites. Timer indicates minutes post cell adherence and LUT: blue-green-yellow from lowest to highest fluorescent signal intensity.

aaw8651_SMaaw8651_SupplementalMaterial_v2