1

Visualizing active viral infection reveals diverse cell fates in synchronized 1

algal bloom demise 2

3

Authors: Flora Vincent1, Uri Sheyn1, Ziv Porat2, Assaf Vardi*1 4

Affiliations: 5

1Department of Plant and Environmental Sciences, Weizmann Institute of Science, Rehovot 6

7610001, Israel. 7

2 Flow Cytometry Unit, Life Sciences Core Facilities, Weizmann Institute of Science, Rehovot 8

7610001, Israel 9

10

*Correspondence to: assaf.vardi@weizmann.ac.il 11

12

Summary: 13

Marine viruses are considered as major evolutionary and biogeochemical drivers of microbial life, 14

through metabolic reprogramming of their host and cell lysis that modulates nutrient cycling1, 15

primary production and carbon export in the oceans2. Despite the fact that viruses are the most 16

abundant biological entities in the marine environment, we still lack mechanistic and quantitative 17

approaches to assess their impact on the marine food webs. Here, we provide the first 18

quantification of active viral infection, during bloom succession of the cosmopolitan 19

coccolithophore Emiliania huxleyi, by subcellular visualization of both virus and host transcripts 20

on a single cell resolution across thousands of cells. Using this novel method, that we coined 21

2

Virocell-FISH, we revealed that distinct transcriptional states co-exist during the infection 22

dynamics, and that viral infection reached only a quarter of the E. huxleyi population although the 23

bloom demised in a synchronized manner. Through a detailed laboratory time-course infection of 24

E. huxleyi by its lytic large virus EhV, we quantitatively show that metabolically active infected 25

cells chronically release viral particles, and that viral-induced lysis is not systematically 26

accompanied by virion increase, thus challenging major assumptions regarding the life cycle of 27

giant lytic viruses. Using Virocell-FISH, we could further assess in a new resolution, the level of 28

viral infection in cell aggregates, a key ecosystem process that can facilitate carbon export to the 29

deep ocean3. We project that our approach can be applied to diverse marine microbial systems, 30

opening a mechanistic dimension to the study of host-pathogen interactions in the ocean. 31

32

One Sentence Summary: Quantifying active viral infection in algal blooms 33

Main Text: 34

Microbes play a crucial role in shaping the marine ecosystem and its biogeochemical cycles4. In 35

particular, abundant photosynthetic micro-eukaryotes and cyanobacteria (phytoplankton) 36

contribute to half of the total primary production on Earth and form the basis of the marine food 37

web5–7. Phytoplankton blooms can undergo synchronized demise following infection by viruses 38

that can reach 108 viruses per L of seawater3,8 and are therefore hypothesized to control the fate of 39

phytoplankton blooms and the subsequent massive release of organic matter that fuels the 40

microbial life via the viral shunt9. One of the main challenges in aquatic virology is to quantify 41

how viruses, through control of the host metabolism and abundance, remodel nutrient fluxes within 42

major biogeochemical cycles and impact the microbial community10,11. This challenge meets the 43

critical need to quantify active viral infection within host cells (virocell), and to assess the 44

3

magnitude of viral-induced mortality within a complex microbial consortium in the marine 45

environment. 46

Approaches such as bulk RNA-seq12,13, ddPCR14, polony15, viral-BONCAT16, phageFISH17 and 47

single cell genomics18,19 have considerably expanded the toolbox of viral ecology and provided 48

insights into viral diversity and viral-encoded auxiliary metabolic genes that remodel the host 49

metabolism20. However, they do not report active viral infection of individual cells thereafter 50

named "virocells", that can be defined as “the living form of the virus in its intra- cellular form”21. 51

Recent advances in single cell transcriptomics have greatly improved our understanding of host-52

virus interactions by uncovering high heterogeneity in the host metabolic states and viral program 53

dynamics within individual virocells22–24. However, these methods lack the ability to visualize 54

infection on the subcellular level that can provide fresh insights into the complex life cycle of 55

viruses in their host cells. Single molecule mRNA Fluorescent In Situ Hybridization (smFISH) 56

enables detection, counting and localization of single mRNA molecules within morphologically 57

intact individual cells25. In this approach, fluorescently labeled viral and host mRNA accumulate 58

in infected cells, whereby probe fluorescence intensity becomes a proxy for gene expression 59

levels26. It therefore provides a new tool to visualize specific microbial interactions within its 60

natural heterogeneous populations. 61

Here, we provide unprecedented quantification of active viral infection and host response of the 62

cosmopolitan unicellular alga Emiliania huxleyi both in the laboratory and in the field. E. huxleyi’s 63

massive oceanic bloom demise are considered to be terminated via infection by its large dsDNA 64

virus EhV3,27,28, a member of the Nucleo-Cytoplasmic Large DNA Viruses29. We detected and 65

characterized host and virus fluorescently labelled mRNA by high throughput imaging flow 66

cytometry and tracked infection dynamics at a single virocell level. Our approach, named Virocell-67

4

FISH, provides major insights into the life cycle of lytic giant virus and its ecology, including how 68

virocells can release virions without lysing, and lyse without releasing virions even though the 69

host undergoes a synchronized demise. 70

To track the metabolic state of E. huxleyi during viral infection, we monitored the mRNA 71

expression of psbA, a chloroplast-encoded gene of the D1 protein. D1 is a major component of the 72

photosystem II (PSII) complex involved in the first step of the light reaction during photosynthesis. 73

D1 has a rapid turnover, requiring constant transcription of psbA30, and is essential for optimal 74

viral infection E. huxleyi31. Concomitantly, we followed the expression of the viral gene mcp, that 75

encodes the major capsid protein expressed at the late stage of the viral program24 (Fig. 1A). 76

Analysis of epifluorescence images of cells at 1 and 24 hours post infection (hpi) showed an 77

increasing fraction of cells expressing viral mcp at 24 hpi in concert with a profound loss of psbA 78

expression (Fig. 1B, 1C). 79

80

In order to quantify active viral infection and host metabolic state in a high throughput manner, a 81

highly resolved time course of E. huxleyi infection at high host:virus ratio was performed. Cell 82

abundance of infected cultures remained relatively stable at 5*105 cell per ml before decreasing at 83

48 hpi during the onset of the lytic phase (Fig. 1D). The first increase in extracellular concentration 84

of virions occurred at 8 hpi before reaching a plateau at 56 hpi of 5.3*108 per mL (Fig. 1E). Based 85

on Virocell-FISH, samples were probed against mcp and psbA, stained with DAPI and acquired 86

using a multispectral imaging flow-cytometer (ImageStream). mcp+ cells represented 19% of the 87

population at 4 hpi, 75% at 24 hpi, plateaued at 60% between 32 and 48 hpi before declining (Fig. 88

1F). Host psbA expression showed that 98% of the non-infected control cells remained psbA+, in 89

contrast with a sharp drop to 12% in the infected cultures between 8 and 24 hpi (Fig. 1F). Infected 90

5

single cells, defined as mcp+, showed large variability in the amounts of mcp mRNA per cell, 91

reflected by a wide dynamic range of intensity values in the mcp probe per cell (from 103 to 105 92

a.u. fluorescence), suggesting cell-cell heterogeneity in viral infection (Fig. 1G). Simultaneously, 93

we detected a dual response in the population with high and low psbA expression (threshold at 1.1 94

x 103 a.u. fluorescence) indicating the loss of photosynthetically active cells in a distinct 95

subpopulation (Fig. 1H). Virocell-FISH therefore enables dual quantification of host and viral 96

genes during infection dynamics across thousands of single cells. 97

6

98

Fig. 1. Virocell-FISH analysis enables high-throughput visualization of active viral infection 99

in single-celled phytoplankton. (A) Simplified workflow of sample processing and data 100

acquisition. After initial fixation, samples are hybridized to custom-made fluorescent probes and 101

are subjected to either epifluorescence microscopy (high resolution) and imaging flow cytometry 102

7

(high throughput) analyses, or both. (B, C) Epifluorescence images of infected E. huxleyi cells at 103

1 and 24 hpi (early and late infection, respectively) in each different channel (Scale bar = 20 µm).. 104

(D, E) Cell and virion concentrations respectively, of infected and non-infected cultures during an 105

E. huxleyi infection at high virus:host ratio. (F) Fraction of mcp+ and psbA+ cells throughout viral 106

infection. Values are presented as the mean ± s.d., n = 3. (G, H) Distribution of mcp and psbA 107

fluorescence intensity values across single cells of E. huxleyi acquired on ImageStream during a 108

time course of viral infection. The dashed line depicts the intensity threshold used to define mcp+ 109

and psbA+ cells according to the fluorescent intensity (103 a.u.). **** P < 0.0001 tested with 110

Linear mixed model fit by REML. T-tests use Satterthwaite's method. 111

112

To investigate E. huxleyi virocell heterogeneity during infection, we plotted host and virus mRNA 113

expressions in parallel and revealed four distinct co-existing subpopulations (Fig. 2A) that we 114

quantified through time of infection (Fig. 2B, Fig. S1). At 0 and 1 hpi, most of the cells were mcp-115

/psbA+ (green gate), indicating photosynthetically active cells. Control cells continuously 116

appeared in this gate. At 4 hpi, 24% of the cells were both mcp+ and psbA+ (red gate), co-117

expressing viral and host genes. Two new subpopulations were clearly distinguishable at 24 hpi: 118

60% of the cells were mcp+/psbA- (yellow gate) as a consequence of psbA transcription shutdown 119

of mcp+/psbA+ cells, and 20% of the cells were mcp-/psbA- (grey gate) and most likely originated 120

from mcp-/psbA+ experiencing gradual photosynthetic shutdown, and cells from mcp+/psbA- that 121

interrupted mcp transcription. We further compared those four subpopulations with single cell dual 122

RNA-Seq data from24 and confirm that our subpopulations correspond to clear stages of viral 123

progression as defined by kinetic classes of viral genes (Fig. S2). Viral release was observed within 124

8 hpi, yet no cell death could be detected by Sytox Green staining (Fig. 2C) which suggested that 125

8

metabolically active infected cells (mcp+/psbA+) are responsible for early viral release. This was 126

supported by the fact that mcp+/psbA+ cells had much higher DAPI intensity than other 127

subpopulations and showed strong positive correlation between mcp and DAPI intensities (Fig. 128

S3), a likely consequence of the virus-induced de novo nucleotide synthesis required to meet the 129

high demand of giant viruses32 (Fig. 2D). Among mcp+/psbA- cells at 24 hpi, mcp and DAPI 130

signals were strongly co-localized, suggesting overlap between viral genome dsDNA and 131

transcripts of viral structural proteins (Fig. S4). psbA intensity in infected cells (mcp+/psbA+) 132

was higher than in non-infected cells (mcp-/psbA+) at 8 and 24 hpi (Fig. S5). Higher psbA 133

expression in mcp+/psbA+ could either reflect enhanced psbA transcription in infected cells or 134

down regulation of psbA translation leading to accumulation of mRNA, as shown for plant 135

viruses33 though transcript and protein levels can be uncorrelated34 and that at the population level 136

cell chlorophyll intensity decreases (Fig. S6). Higher psbA in mcp+ population emphasized the 137

interdependence between optimality of infection (mcp expression) and host metabolic regulation 138

(psbA expression). 139

Interestingly, based on our intracellular virocell infection dynamics, we quantify fundamental 140

parameters in the life cycle of a marine giant viruses. We estimate that the eclipse phase, in which 141

viruses have penetrated the host but no virions are produced35, occurs within the first 4 hours, 142

based on the induction of mcp+/psbA+ cells, without virion increase (Fig. 2C). The onset of the 143

maturation phase, during which viral progeny is released, occurs within 8 hours. The distinct 144

infection states can be associated with unique cell fates: metabolically active infected cells 145

(mcp+/psbA+) are responsible for early viral release most likely through budding whilst 146

mcp+/psbA- likely contribute less to this process. The mcp-/psbA- subpopulation, correlated with 147

Sytox Green staining (Fig. S1B), suggests that a substantial fraction of the cells lyse without 148

9

releasing viruses after 56 hpi. These cells can die from accumulation of DNA damage36, cytotoxic 149

viral glycosphingolipids coming from neighboring infected cells37, or by abortive infection. The 150

accumulation of these cells during lytic phase can originate from either the mcp+/psbA+ 151

subpopulation that halts mcp transcription before lysis, or from the mcp-/psbA+ subpopulation that 152

shuts down psbA transcription24. At later time point, the mcp-/psbA+ cells might serve as the seed 153

for the resistant phenotypes often observed in small subpopulation that recovered from viral 154

infection38. The transition between different phenotypic states could be mediated by an array of 155

infochemicals, such as viral glycosphingolipids37,39, or by extracellular vesicles40 (Fig. 2E). Virion 156

egress from the host cells without induction of cell death may provide an optimal co-existence 157

mechanism during bloom initiation phase41. 158

159

Fig. 2. Heterogeneity in transcriptional states during viral infection reveals distinct potential 160

cell fates. (A) The central panel represents co-expression of host psbA and viral mcp (X and Y 161

axes, respectively in arbitrary units of fluorescence) in an infected E. huxleyi culture 24 hpi. Four 162

subpopulations, representing distinct transcriptional states, are observed: mcp-/psbA+ (green gate), 163

mcp+/psbA+ (red gate), mcp+/psbA- (yellow gate) and mcp-/psbA- (grey gate). Single-cell 164

10

morphology and fluorescent signal of each subpopulation is presented, showing brightfield image 165

(BF), mcp (yellow), DAPI (purple) and psbA (red). (B) Relative abundance of the four 166

subpopulations throughout the course of infection. (C) Close-up on virion production (virions ml-167

1), cell death (Sytox+), and mcp+ dynamics in the 24 first hours post infection. (D) Comparison of 168

DNA content based on DAPI intensity between the four subpopulations (with more than 25 events 169

per population) throughout infection. Comparison of DAPI intensity was with a Linear mixed 170

model fit by REML. T-tests use Satterthwaite's method. * P < 0.05, **** P < 0.0001. (E) 171

Schematic model of potential cell fates of the different subpopulations. 172

173

The consequences of viral infection in the ocean are not limited to the cell fate of the individual 174

host cell. Viral infection can also lead to aggregate formation, mediated by production of 175

transparent exopolymer particles (TEP) consisting of acidic polysaccharides, which are classically 176

quantified in bulk by staining with Alcian Blue42. Aggregates accelerate sinking of biomass, 177

contribute to “marine snow” and enhance carbon export to the deep sea3. In order to assess the 178

contribution of viral infection to aggregate formation, we quantified aggregated E. huxleyi cells 179

and characterized the infection state among individual cells within the aggregates using Virocell-180

FISH (Fig. 3A). Upon infection, up to 40% of the algal cells (DAPI positive fraction) were found 181

within aggregates (Fig. 3B). At the early stage of infection, aggregates are rare (<10%), mostly 182

infected (75% of the aggregates were mcp+) and photosynthetically active (100% were psbA+) 183

(Fig. 3D, 3E). At the onset of the lytic phase (48 hpi), aggregates were highly abundant, less 184

infected (25% were mcp+) and less photosynthetically active (30% were psbA+). In infected 185

cultures at 8 hpi, the fraction of mcp+ aggregates was higher than the fraction of mcp+ single cells 186

(75% and 45%, respectively, Fig. 1F and Fig. 3D). We hypothesize that aggregates enriched with 187

11

infected cells may favor sinking of newly produced virions out of the photic zone and prevent 188

further dissemination of viruses in the bloom, suggesting a host defense strategy3,28. The overall 189

presence of infected cells in aggregates corroborates previous observations of TEP-attached 190

viruses42,43. This could further explain the estimates of 109 viruses per gram of marine sediment, 191

some being infectious, hence may serve as a possible inoculum to infect subsequent blooms44,45. 192

193

Fig. 3. Quantification of viral infection within aggregates formed during infection dynamics. 194

(A) Based on the DAPI mask area and circularity (the degree of the mask’s deviation from a circle), 195

three populations are identified as single cells, doublets and aggregates (aggregates are defined by 196

a DAPI area above 60 µm2). (B) Imaging flow-cytometry images of non-infected (top) and infected 197

(bottom) aggregates in the brightfield (BF), mcp (yellow), Side Scatter (SSC as a proxy for 198

calcification, pink), DAPI (blue), and psbA (red) channels. (C) Fraction of aggregates (DAPI area 199

above 60 µm2) out of all DAPI+ events in infected and non-infected cultures. (D, E). 200

Quantification of mcp+ and psbA+ aggregates respectively, through time across three replicates 201

between infected and non-infected cultures. **** P < 0.0001 (t-test). 202

12

203

One of the fundamental knowledge gaps in aquatic virology is the ability to quantify active viral 204

infection in natural populations and to mechanistically resolve the viral impact on microbial 205

ecosystems. We therefore conducted a mesocosm experiment in a Fjord near Bergen, Norway, and 206

collected samples for Virocell-FISH analysis across different phases of an E. huxleyi bloom. We 207

quantified viral infection specifically in E. huxleyi cells by using a 28S ribosomal probe unique to 208

that algae, a psbA probe as a proxy for host metabolic state and a probe targeting viral mcp 209

transcripts (Fig. 4A). Virocell-FISH analysis conducted by imaging flow cytometry successfully 210

revealed the fraction of infected E. huxleyi cells during bloom succession, based on over 5000 cells 211

at each time point (Fig. S7). From the initiation of the bloom to the peak in E. huxleyi abundance, 212

infection level (% of mcp+ cells) was minimal (<1%). A tipping point was detected in which the 213

fraction of infection cells rose from 1.51% on day 17 to 23.22% on day 18 while cell abundance 214

dropped at the onset of bloom demise (Fig. 4B). Strikingly, the maximal fraction of infected cells 215

on the demise phase, did not reach beyond 27% of E. huxleyi population even at the highest levels 216

of E. huxleyi infection. Furthermore, mcp+ cells displayed higher DAPI intensity than mcp- cells 217

(Fig. 3C), confirming laboratory results regarding the induction of de-novo nucleotide synthesis 218

in infected cells. Quantification of co-expression of mcp and psbA over the course of the bloom 219

(Fig. S8) revealed a major drop in mcp-/psbA+ cells between days 16 and 18, accompanied by an 220

increase in mcp+/psbA- cells. From day 19 onwards a large fraction of the E. huxleyi population 221

was mcp-/psbA- resembling increase in the fraction of cell death during bloom demise (Fig. 4F). 222

These results suggest algal bloom demise can be synchronized despite co-existence of diverse 223

virocell states and only a subpopulation of infected cells. We suggest that other mortality agents 224

might act in concert with viral lysis to control synchronized bloom demise at a later stage. These 225

13

top-down regulators may include pathogenic bacteria46, eukaryotic parasites and grazers3,47,48, 226

programmed cell death of non-infected bystander cells mediated by release of viral-derived 227

infochemicals37,39 or extracellular vesicles40. Alternatively, a highly synchronized viral infection 228

and release might have occurred in a large fraction of the E. huxleyi population in a narrow time 229

frame between day 17 and 18. Indeed, recent reports demonstrated that viruses are intrinsically 230

synchronized with the daily rhythms of their photosynthetic host both in controlled lab experiments 231

and in natural populations49,50. 232

233

Fig. 4. Visualization of active viral infection during a natural E. huxleyi bloom. (A) Applying 234

Virocell-FISH to track viral infection of E. huxleyi cells within mixed natural microbial 235

populations. Viral mRNA was targeted using the mcp probe whilst host mRNA was monitored 236

using psbA; an additional probe specifically targeting the 28S ribosomal RNA of E. huxleyi was 237

used to identify our host of interest, based on the intensity and max pixel of the probe signal. (B) 238

Fraction of infected (mcp+) E. huxleyi cells (orange), abundance of calcified E. huxleyi cells 239

(black) and EhV like particles (brown) throughout bloom succession (days post inoculation) in a 240

14

mesocosm experiment. Active viral infection in E. huxleyi cells was measured using Virocell-241

FISH; E. huxleyi and EhV-viral like particles abundance were measured by flow-cytometry. Days 242

with 0.5 digit represent evening samples. (C) Comparison of DAPI intensities between infected 243

and non-infected E. huxleyi populations throughout bloom succession. (D, E) Visualizing 244

subpopulation of E. huxleyi single cells and aggregates respectively, using ImageStream. (F) 245

Quantification of the relative abundance of the four subpopulations in single cells throughout the 246

bloom succession, by co-probing of host and viral mRNA (psbA and mcp respectively). 247

248

Our approach opens up new avenues to study individual virocells in the marine environment, by 249

following infection dynamics both intra- and extracellularly of morphologically intact cells. It 250

unravels fundamental aspects of the life cycle of a large virus by linking scales between single cell 251

and population level dynamics, showing viral increase without cell lysis, and cell lysis without 252

viral increase. Such high-resolution examination of host and virus transcriptional states reveals 253

cell-to-cell heterogeneity in viral infection, mediated by the dependence on the host metabolic 254

state, despite synchronized bloom demise. Concomitantly, it can inform about viral hijacking 255

mechanisms of the host photosynthetic machinery. Coupling enhanced viral and photosynthetic 256

transcriptions may require to maintain photosynthetic activity to energize the redirection of carbon 257

flux from the Calvin cycle to the pentose phosphate pathway in order to enhance production of 258

NADPH required for de novo DNA synthesis51 to meet the high demand of DNA synthesis of giant 259

viruses and to enhance the antioxidant capacity under oxidative stress52. Therefore, we propose 260

that a functional convergence arises among viruses infecting phototrophs from diverse 261

evolutionary origin, whereby cyanophages53, small RNA plant viruses54 and large algal dsDNA 262

viruses share the strategy to manipulate their photosynthetic host metabolism by different 263

15

mechanisms. Quantifying the virocells infection and metabolic states, through their diverse 264

phenotypes will help integrate viral-dependent processes in ecosystem models by providing 265

necessary numbers of infection dynamics, such as the fraction of actively infected cells, resistant 266

cells, the nature of sinking biomass, as well as their genetic composition as a function of time and 267

other stresses55. Finally, this quantitative approach can be expanded to diverse host-pathogen and 268

host-symbiont systems with important ecological significance, adding a mechanistic dimension to 269

the field of marine microbial ecology. 270

271

References and Notes: 272

273

1. Zimmerman, A. E. et al. Metabolic and biogeochemical consequences of viral infection in 274

aquatic ecosystems. Nat. Rev. Microbiol. 18, 21–34 (2020). 275

2. Suttle, C. A. Marine viruses — major players in the global ecosystem. Nat. Rev. 276

Microbiol. 5, 801–812 (2007). 277

3. Laber, C. P. et al. Coccolithovirus facilitation of carbon export in the North Atlantic. Nat. 278

Microbiol. 3, 537–547 (2018). 279

4. Azam, F. Microbial control of oceanic carbon flux: The plot thickens. Science 280, 694–280

696 (1998). 281

5. Fenchel, T. Marine plankton food chains introduction: the classical view of plankton food 282

chains. Ann. Rev. Ecol. Syst 19, 19–38 (1988). 283

6. Field, C. B. Primary Production of the Biosphere: Integrating Terrestrial and Oceanic 284

Components. Science 281, 237–240 (1998). 285

16

7. Bar-On, Y. M., Phillips, R. & Milo, R. The biomass distribution on Earth. Proc. Natl. 286

Acad. Sci. U. S. A. 115, 6506–6511 (2018). 287

8. Lehahn, Y. et al. Decoupling physical from biological processes to assess the impact of 288

viruses on a mesoscale algal bloom. Curr. Biol. 24, 2041–2046 (2014). 289

9. Wilhelm, S. W. & Suttle, C. A. Viruses and Nutrient Cycles in the Sea. Bioscience 49, 290

781–788 (1999). 291

10. Weitz, J. S. & Wilhelm, S. W. Ocean viruses and their effects on microbial communities 292

and biogeochemical cycles. F1000 Biol. Rep. 4, (2012). 293

11. Brum, J. R. et al. Patterns and ecological drivers of ocean viral communities. Science 348, 294

6237 (2015). 295

12. Sieradzki, E. T., Ignacio-Espinoza, J. C., Needham, D. M., Fichot, E. B. & Fuhrman, J. A. 296

Dynamic marine viral infections and major contribution to photosynthetic processes 297

shown by spatiotemporal picoplankton metatranscriptomes. Nat. Commun. 10, 1169 298

(2019). 299

13. Moniruzzaman, M. et al. Virus-host relationships of marine single-celled eukaryotes 300

resolved from metatranscriptomics. Nat. Commun. 8, 16054 (2017). 301

14. Tadmor, A. D., Ottesen, E. A., Leadbetter, J. R. & Phillips, R. Probing individual 302

environmental bacteria for viruses by using microfluidic digital PCR. Science 333, 58–62 303

(2011). 304

15. Baran, N., Goldin, S., Maidanik, I. & Lindell, D. Quantification of diverse virus 305

populations in the environment using the polony method. Nat. Microbiol. 3 62-72 (2017). 306

16. Pasulka, A. L. et al. Interrogating marine virus-host interactions and elemental transfer 307

17

with BONCAT and nanoSIMS-based methods. Environ. Microbiol. 20, 671-692. (2018). 308

17. Allers, E. et al. Single-cell and population level viral infection dynamics revealed by 309

phageFISH, a method to visualize intracellular and free viruses. Environ. Microbiol. 15, 310

2306–2318 (2013). 311

18. Labonté, J. M. et al. Single-cell genomics-based analysis of virus-host interactions in 312

marine surface bacterioplankton. ISME J. 9, 2386–2399 (2015). 313

19. Needham, D. M. et al. A distinct lineage of giant viruses brings a rhodopsin photosystem 314

to unicellular marine predators. Proc. Natl. Acad. Sci. U. S. A. 116, 20574–20583 (2019). 315

20. Hurwitz, B. L., Hallam, S. J. & Sullivan, M. B. Metabolic reprogramming by viruses in 316

the sunlit and dark ocean. Genome Biol. 14, R123 (2013). 317

21. Forterre, P. The virocell concept and environmental microbiology. ISME J. 7, 233–236 318

(2013). 319

22. Cristinelli, S. & Ciuffi, A. The use of single-cell RNA-Seq to understand virus–host 320

interactions. Curr. Opin. in Virol. 29, 39–50 (2018). 321

23. Rosenwasser, S. et al. Unmasking cellular response of a bloom-forming alga to viral 322

infection by resolving expression profiles at a single-cell level. PLoS Pathog. 15, (2019). 323

24. Ku, C. et al. A single-cell view on alga-virus interactions reveals sequential transcriptional 324

programs and infection states. Sci. Adv. 6, (2020). 325

25. Raj, A., van den Bogaard, P., Rifkin, S. A., van Oudenaarden, A. & Tyagi, S. Imaging 326

individual mRNA molecules using multiple singly labeled probes. Nat. Methods 5, 877–327

879 (2008). 328

18

26. Skinner, S. O., Sepúlveda, L. A., Xu, H. & Golding, I. Measuring mRNA copy number in 329

individual Escherichia coli cells using single-molecule fluorescent in situ hybridization. 330

Nat. Protoc. 8, 1100–1113 (2013). 331

27. Wilson, W. H. et al. Isolation of viruses responsible for the demise of an Emiliania 332

huxleyi bloom in the English Channel. J. Mar. Biol. Assoc. 82, 369–377 (2002). 333

28. Sheyn, U. et al. Expression profiling of host and virus during a coccolithophore bloom 334

provides insights into the role of viral infection in promoting carbon export. ISME J. 12, 335

704–713 (2018). 336

29. Koonin, E. V. & Yutin, N. Origin and evolution of eukaryotic large nucleo-cytoplasmic 337

DNA viruses. Intervirology 53, 284–292 (2010). 338

30. Mattoo, A. K., Marder, J. B. & Edelman, M. Dynamics of the photosystem II reaction 339

center. Cell 56, 241–246 (1989). 340

31. Thamatrakoln, K. et al. Light regulation of coccolithophore host-virus interactions. New 341

Phytol. 221, 1289–1302 (2019). 342

32. Rosenwasser, S., Ziv, C., Creveld, S. G. Van & Vardi, A. Virocell Metabolism : Metabolic 343

Innovations During Host – Virus Interactions in the Ocean. Trends Microbiol. 24, 1–12 344

(2016). 345

33. Lehto, K., Tikkanen, M., Hiriart, J. B., Paakkarinen, V. & Aro, E. M. Depletion of the 346

Photosystem II Core Complex in Mature Tobacco Leaves Infected by the Flavum Strain 347

of Tobacco mosaic virus. Mol. Plant-Microbe Interact. 16, 1135–1144 (2003). 348

34. Taniguchi, Y. et al. Quantifying E. coli proteome and transcriptome with single-molecule 349

sensitivity in single cells. Science 329, 533–538 (2010). 350

19

35. Flint, S. J., Racaniello, V. R., Enquist, L. W. & Skalka, A. M. Principles of virology, 351

Volume 1: molecular biology. Princ. Virol. Vol. 1 Mol. Biol. (2008). 352

36. Bidle, K. D., Haramaty, L., Ramos, J. B. & Falkowski, P. Viral activation and recruitment 353

of metacaspases in the unicellular coccolithophore, Emiliania huxleyi. Proc. Natl. Acad. 354

Sci. U. S. A. 104, 6049–6054 (2007). 355

37. Vardi, A. et al. Viral Glycosphingolipids Induce Lytic Infection and Cell Death in Marine. 356

Science 326, 861 (2009). 357

38. Frada, M. J. & Vardi, A. Morphological switch to a resistant subpopulation in response to 358

viral infection in a bloom-forming marine microalgae. PLOS Pathog. 1–17 (2017). 359

39. Schleyer, G. et al. In plaque-mass spectrometry imaging of a bloom-forming alga during 360

viral infection reveals a metabolic shift towards odd-chain fatty acid lipids. Nat. 361

Microbiol. 4, 527–538 (2019). 362

40. Schatz, D. et al. Communication via extracellular vesicles enhances viral infection of a 363

cosmopolitan alga. Nat. Microbiol. 2, 1485–1492 (2017). 364

41. Yau, S. et al. Virus-host coexistence in phytoplankton through the genomic lens. Sci. Adv. 365

6, (2020). 366

42. Vardi, A. et al. Host – virus dynamics and subcellular controls of cell fate in a natural 367

coccolithophore population. Proc. Natl. Acad. Sci. U. S. A. 114, 11446-11451 109, 19327–368

19332 (2012). 369

43. Mari, X., Kerros, M. E. & Weinbauer, M. G. Virus attachment to transparent 370

exopolymeric particles along trophic gradients in the southwestern lagoon of New 371

Caledonia. Appl. Environ. Microbiol. 73, 5245–5252 (2007). 372

20

44. Coolen, M. J. L. 7000 years of Emiliania huxleyi viruses in the Black Sea. Science 333, 373

451–452 (2011). 374

45. Lawrence, J. E., Chan, A. M. & Suttle, C. A. Viruses causing lysis of the toxic bloom-375

forming alga Heterosigma akashiwo (Raphidophyceae) are widespread in coastal 376

sediments of British Columbia, Canada. Limnol. Oceanogr. 47, 545–550 (2002). 377

46. Barak-Gavish, N. et al. Bacterial virulence against an oceanic bloom-forming 378

phytoplankter is mediated by algal DMSP. Sci. Adv. 4, (2018). 379

47. Chambouvet, A. et al. Diverse molecular signatures for ribosomally ‘active’ Perkinsea in 380

marine sediments. BMC Microbiol. 14, 110 (2014). 381

48. Mayers, K. M. J. et al. Growth and mortality of coccolithophores during spring in a 382

temperate Shelf Sea (Celtic Sea, April 2015). Prog. Oceanogr. 177, 101928 (2019). 383

49. Liu, R., Liu, Y., Chen, Y., Zhan, Y. & Zeng, Q. Cyanobacterial viruses exhibit diurnal 384

rhythms during infection. Proc. Natl. Acad. Sci. U. S. A. 116, 14077–14082 (2019). 385

50. Aylward, F. O. et al. Diel cycling and long-term persistence of viruses in the ocean’s 386

euphotic zone. Proc. Natl. Acad. Sci. U. S. A. 114, 11446-11451 (2017). 387

51. Puxty, R. J., Millard, A. D., Evans, D. J. & Scanlan, D. J. Viruses Inhibit CO 2 Fixation in 388

the Most Abundant Phototrophs on Earth. Curr. Biol. 26, 1585–1589 (2016). 389

52. Sheyn, U., Rosenwasser, S., Ben-Dor, S., Porat, Z. & Vardi, A. Modulation of host ROS 390

metabolism is essential for viral infection of a bloom-forming coccolithophore in the 391

ocean. ISME J. 10, 1742–1754 (2016). 392

53. Thompson, L. R. et al. Phage auxiliary metabolic genes and the redirection of 393

cyanobacterial host carbon metabolism. Proc. Natl. Acad. Sci. U. S. A. 108, (2011). 394

21

54. Zhao, J., Zhang, X., Hong, Y. & Liu, Y. Chloroplast in plant-virus interaction. Front. in 395

Microbiol. 7, (2016). 396

55. Record, N. R., Talmy, D. & Våge, S. Quantifying tradeoffs for marine viruses. Front. 397

Mar. Sci. 3, (2016). 398

399

400

Acknowledgments: We thank S. Itzkovitz, L. Farack, S. Ben-Moshe for guidance in smFISH 401

protocols, analysis, and fruitful discussions; D. Hirsch for training on the epifluorescence 402

microscope. The AQUACOSM-VIMS consortium for the mesocosm experiment and all Vardi lab 403

members for discussion. We particularly thank G. Schleyer and D. Schatz for critical reading, and 404

R. Avraham for useful critical feedback; Funding: This work was supported by the European 405

Research Council (ERC) CoG (VIROCELLSPHERE grant # 681715; A.V.), the Israeli Academy 406

for Science and Humanities and Weizmann Dean of Faculty (F.V.); Author contributions: F.V. 407

and A.V. designed and conceptualized the study. U.S. developed smFISH on E. huxleyi and F.V. 408

implemented it in high throughput. Z.P. helped with ImageStream analysis F.V. prepared the cell 409

cultures, performed all acquisition and data analysis. F.V. prepared the figures and wrote the 410

manuscript with A.V. and input from U.S.; A.V. supervised the study; Competing interests: 411

Authors declare no competing interests; Data and materials availability: All data, code, and 412

materials used in the analysis are available upon demand. 413

Supplementary Materials: 414

Materials and Methods 415

Table S1 416

22

Table S1. Calculation of virus:host ratio using the MPN method 417

Figures S1-S13 418

Fig. S1. Dynamics of subpopulations during infection in EhV201 419

Fig. S2. Comparison of Virocell-FISH with single-cell RNA-Seq 420

Fig. S3. mcp and DAPI intensities in single cells throughout infection 421

Fig. S4. Morphology of metabolically active infected and non-infected cells 422

Fig. S5. Temporal dynamics of psbA intensities in psbA+ subpopulations 423

Fig. S6. Chlorophyll mean intensity during infection 424

Fig. S7. Adapting Virocell-FISH to target specifically E.huxleyi in environmental samples 425

Fig. S8. Defining mcp/psbA subpopulations in environmental samples. 426

Fig. S9. Pipeline of single cell identification in ISX (Only SM) 427

Fig. S10. Comparison of the Virocell-FISH method with or without RNAase treatment (Only SM) 428

Fig. S11. Quantifying infected cells in microscopy data using CellProfiler (Only SM) 429

Fig. S12. Gene expression of psbA and mcp in single cell RNA-seq sequencing data (Only SM) 430

Fig. S13. Flow cytometry gates used for E. huxleyi cells in the field (Only SM) 431