Skip to the content

Sunscreen Products Increase Virus Production Through Prophage Induction in Marine Bacterioplankton

Sunscreen Products Increase Virus Production Through Prophage Induction in Marine Bacterioplankton

  1. Danovaro, C. Corinaldesi

Institute of Marine Science, University of Ancona, Via Brecce Bianche, 60131 Ancona, Italy Received: 10 June 2002; Accepted: 25 October 2002; Online publication: 28 January 2003



Classical pollutants (e.g., hydrocarbon, pesticides) have been recently recognized to induce lytic cycle in lysogenic bacteria, but information on micro-pollutants is almost completely lacking. We investigated the effects of cosmetic sun products (sunscreen and solar oil) on viral abun- dance and bacterial activity. We found that both sunscreen and solar oil acted as pollutants, inducing viral development and controlling bacterial abundance and production, thus leading to an increase of the virus to bacterium ratio. Short-term experiments revealed that sunscreen supplementation induced the lytic cycle in a large fraction of total bacterial abundance (13–24% of bacteria, at low and high concentrations, respectively), whereas solar oil had a lower impact (6–9%). A synchronized development of the phage–host system was observed only after sun- screen addition. The addition of sunscreen, even at low concentrations, had a significant impact on all enzymatic activities (aminopeptidase, glucosidase, and phosphatase), which increased significantly. However, when enzymatic activities were normalized per cell, a selective en- hancement was observed for certain enzymes (e.g., aminopeptidase) and inhibition for others (e.g., glucosidase). These results indicate that sunscreen products can modify C, N, and P biogeochemical cycling in seawater and increase virus abundance through prophage induction in marine bacterioplankton.








During the past three decades, the impact of chemical pollution in the marine environment has focused almost exclusively on the conventional ‘‘priority’’ pollutants (es- pecially persistent toxic/carcinogenic pesticides and in- dustrial intermediates). Comparatively, bioactive


Correspondence to: R. Danovaro; E-mail:


chemicals received much less attention. Production, con- sumption, and the consequent release into the environ- ment of active ingredients in personal care products (e.g., chemical fragrances and perfumes, cosmetic and sun- screen agents) are largely increasing. These compounds can be introduced directly into the aquatic environment as complex mixtures, and/or indirectly through domestic sewage and treatment plants [18]. Recent studies proved that these chemicals and their degradation products might


have a significant impact on aquatic organisms, even at very  low  concentrations  (i.e.,  at  ng–mg  L)1  at  levels  at which they might occur in nature [18]).

Production and consumption of cosmetic sun products is reaching unexpected levels (every person uses, on av- erage, 65–130 mg of sunscreen per application [12]), thus assuming a potentially important role in environmental contamination. Annually, 7900 tons of sunscreen products are consumed in Germany, and it can be estimated that 20,000 tons are released annually in the northern Medi- terranean.

Sunscreens fall into two broad categories: chemical and physical sun-blocking agents. The two categories are often


phage induction in natural populations of marine bacteria, with a high percent efficiency of induction (up to 75% [11, 26, 27]). Prophage induction of marine bacteria might have important ecological and biogeochemical implications [21, 40, 42], but the effect of solar products and sun-blocking compounds is still to be assessed.

We investigated the effects of sunscreen products (i.e., both sunscreen and solar oil), at different concentrations, on viral and bacterial natural assemblages in order to as- sess the impact on microbial loop functioning and po- tential biogeochemical consequences of the introduction of these micropollutants at sea.


used in combination to offer a full protection against UV-                                                                                     


A (400–315 nm) and UV-B (315–280 nm) radiation. Chemical filters include benzophenones, PABA and PABA esters, cinnamates, salicylates, camphor derivatives, dib- enzoylmethanes, and anthranilates, in concentration up to 10%. Physical filters include talc, kaolin, zinc oxide, and titanium dioxide (in concentration up to 25%). Sunscreen products contain a cosmetic lipid base and other sub- stances (vitamins, b-carotene, alpha-bisabolol, aloe, hyal- uronic acid, panthenol, vegetable extracts). In addition, sunscreens contain potentially harmful components that might have an ecological impact: (i) sun-blocking com- pounds (e.g., methylbenzylidene camphor), that being li- pophilic, bioaccumulate in different fish species; (ii) antimicrobial preservatives (such as the parabens, alkyl-p- hydroxybenzoates), whose toxicity is assumed to be low, but the continual release of these benzoates into the en- vironment can become harmful for  aquatic  organisms  [33]; (iii) retinoids (low-molecular-weight lipophilic de- rivatives of vitamin A), which have impact on embryonic development [28]; (iv) imidazolidinyl urea (another pre- servative) that inhibits protein and DNA synthesis in sea urchin eggs [2]; (v) EDTA (a water-soluble chelating  agent) that affects metabolic pathways [7]; (vi) aromatic hydrocarbons (benzene, benzophenones, butylated hy- droxyanisole and butylated hydroxytoluene), benzoic acid, and PABA (para-aminobenzoic acid), which have been demonstrated to be harmful to aquatic organisms (Int. Chemical Safety Cards, html).

Information on viral and bacterial response to pollutants in the marine environment is almost completely lacking. Recently, it has been demonstrated that, in different envi- ronments, Aroclor 1248, a PCB mixture, a pesticide mixture, and hydrocarbons (bunker C fuel oil #6) determined pro-


Materials and Methods

Cosmetic Sun Products


In this study, after screening several commercial sun products, we selected a sunscreen containing both chemical and physical sun-blocking agents (Ambre solaire Mexoryl sx 7, Laboratoires Garnier) and a solar oil without UV protection (Bilboa Sun Oil, Cadey). Both products displayed widespread utilization and their composition included all substances mentioned above. The use of two different solar products (with and without UV protection agents) at two different concentrations (versus nontreated con- trols) allowed identification, by comparison, of the effects of the solar products with different characteristics (and micropollutant concentrations) on investigated microbial parameters.


Experimental System


Surface seawater samples were collected in October 2000 at Portonovo (Ancona, northern Adriatic Sea). The northern Adriatic is one of the most eutrophic areas of the entire Medi- terranean, and it is characterized by high viral abundance and high values of the ratio of viruses to bacteria, which have been hypothesized to reflect eutrophication gradients present in this area [36]. After sampling, seawater samples were immediately filtered through 200-mm pore-size filters and then through 2-mm pore-size filters to remove suspended particles, nanoflagellates, phytoplankton, and larger organisms (the removal of protozoa and micro-algae has also been subsequently checked under epi- fluorescence microscopy [16]). Fifteen L filtered seawater was transferred into 15 (1-L) polyethylene Whirlpack bags and in- cubated in the dark (in order to minimize the effects due to virus decay induced by solar radiation and to avoid the possible in- terference due to picophytoplankton autotrophic production)  at

in situ temperature (i.e., 15°C). This experimental design, as

explicated by Bratbak et al. [8], is likely to minimize or even exclude the presence of a significant ‘‘bottle effect,’’ when com- pared to processes  occurring in natural seawater samples. After   24 h, the polyethylene microcosms were divided into five groups. Each group consisted  of three  independent microcosms  and   was


subjected to a different treatment): group 1, control (untreated microcosms); group 2, ‘‘Sunscreen’’ (microcosms supplemented with  50  mL  L)1  of  sunscreen);  group  3,  ‘‘Sunscreen  max’’  (mi- crocosms supplemented with 500 mL L)1 of sunscreen); Group 4, ‘‘Oil’’  (microcosms  supplemented  with  50  mL  L)1  of  solar  oil); and group 5, ‘‘Oil max’’ (microcosms supplemented with 500 mL L)1 of solar oil). Time of preincubation and the duration of the experiments were set up according to previous studies conducted on polyethylene microcosms [38, 39]. Group 1, 3, and 5 were also used for the short-term experiment.


Long-term and short-term experiments


The long-term experiment lasted 90 days: t0, t1 (4 h after inoc- ulation of sunscreen products), t2 (1 day), t3 (2 d), t4 (3 d), t5 (5 d), t6 (8 d), t7 (15 d), t8 (31 d), t9 (90 d). At each sampling time, three replicate subsamples (1–5 mL) were collected from each bag for the analysis of the following parameters: viral and bac- terial abundance, bacterial biomass, bacterial carbon production, b-D-glucosidase, L-aminopeptidase, and phosphatase activities. In order to reduce possible artifacts [8] all microcosms were gently mixed (for 30 s) daily.

The short-term experiment was set up after 15 days (to monitor and compare viral dynamics of different treatments after the sunscreen effects were stabilized [8]) and lasted 2 h during which subsamples were taken every 6 min from control, ‘‘Sun- screen max’’ and ‘‘Oil max’’ microcosms (three microcosms for each treatment and for the control), utilized for the long-term experiment. Subsamples were analyzed only for the determina- tion of viral and bacterial abundance.


Organic Content and Biochemical Composition on Sunscreen and Solar Oil

Total organic matter was determined as the difference between dry weight (60°C, 24 h) of the sunscreen and weight of the residue after combustion (i.e., after complete oxidation of organic matter to CO2) at 450°C (for 2 h [32]). Total proteins were determined according to the method of Hartree [23] that is based on the protein property to react with copper tartrate in a basic envi- ronment. Concentrations are reported as mg albumin equivalents in 50 and 500 mL of product. Total carbohydrates were analyzed, using the phenol–sulfuric method, according to the method of Dubois et al. [19] and concentrations are expressed as mg glucose equivalents in 50 and 500 mL of product. Lipids were extracted by direct elution with chloroform and methanol and determined after carbonization in sulfuric acid at 180°C [6, 29]. Lipid con- centrations are reported as mg tripalmitine equivalents in 50 and 500 mL of product. Organic P was determined as the difference between total P and inorganic P [3, 35].



Viral and Bacterial Parameters


Direct counts of bacteria and viruses were carried out using the method described by Noble and Fuhrman [31] with few modifi-


cations [13, 15, 17]. Subsamples were preserved in 0.02 mM prefiltered formalin (2% final concentration). Subsamples (100 mL were diluted 1:10 in Milli-Q prefiltered, sterile water and then concentrated onto 0.02 mM pore size filters (Anodisc 25 mM, Al2O3), and colored with 20 mL SYBR Green I (stock solution diluted 1:20). Filters were incubated in the dark for 15 min and mounted on glass slides with a drop of 50% phosphate-buffered saline and 50% glycerol, containing 0.5% ascorbic acid. Viral and bacterial counts were carried out by epifluorescence microscopy on at least 10 fields per slide in order to count at least 200 cells per replicate. SYBR Green I has been repeatedly utilized in eco- logical studies [41]. Recent studies reported that this stain, since it fades quickly, can underestimate virus counting when large virus numbers are encountered (i.e., <1011 L)1 [5]. This was not the case in our study so, therefore, we assumed that virus counting was not underestimated.

Bacterial biovolume was estimated assigning each cell to a dimensional class, according to bacterial length and shape. Bacterial biovolume was converted to bacterial biomass assum- ing an organic carbon content of 310 fgC mM)3 [20].

Bacterial carbon production was determined by 3H-leucine incorporation [34] (final concentration 10 nM, incubated for 1  h at in situ temperature). Incubations were terminated with ethanol 80%; subsamples were washed with TCA and ethanol 80% and 1 mL scintillation liquid was added to the supernatant for the determination of incorporated radioactivity (DPM by scintillation).



Enzymatic Activities


Extracellular enzyme activities (L-aminopeptidase, b-D-glucosi- dase, and phosphatase) were determined by cleavage of artificial fluorogenic substrates [10, 25]. For the determination of b-D- glucosidase, L-aminopeptidase, and phosphatase activities, MFU- b-glucopiranoside (MUF-glu, final concentration 200 mM), L- leucina-4-methylcoumarinile-7-amide (Leu-MCA, final concen- tration 200 mM), and 4-methylumbelliferone phosphate (MUF-P, final concentration 50 mM) were utilized, respectively. Incuba- tions were performed in the dark at in situ temperature for 1 h. Fluorescence caused by the enzymatic cleavage of artificial sub- strates was measured fluorometrically (365 nM excitation and 455 nM emission for b-D-glucosidase and phosphatase, 440 nM and 380 nM, respectively, for aminopeptidase [13]). All analyses were carried out in three replicates from each of the three rep- licate microcosms and for each treatment (i.e., control, ‘‘Sun- screen,’’ ‘‘Sunscreen max,’’ ‘‘Oil,’’ and ‘‘Oil max’’).



Statistical Analysis


A two-way ANOVA was utilized to investigate differences be- tween treatments and control for all of microbiological parame- ters (with time and treatment as source of variability). When differences were statistically significant a post-hoc Tukey’s test was performed.
















Fig. 1. Long-term changes in  viral abundance in the control, ‘‘Sunscreen’’ (SS), and ‘‘Sunscreen max’’ (SSmax) microcosms (a); and viral abundance in  the  control, ‘‘Oil,’’ and ‘‘Oil max’’ microcosms (b).

Standard deviations (each value represents 3 replicate analyses of 3 microcosms) are reported at all sampling times (not visible when bars are within the mark).





Biochemical Analyses


the control by a factor of 2–4. Highest viral abundance was observed immediately after the beginning of the experi- ment  in  the  ‘‘Oil’’  (10.4  · 109  viruses  L)1 at  t0)  and  ‘‘Oil

)1           max’’ microcosms (14.6 · 10 viruses L · 109 viruses L)1 at


The sunscreen inoculated (final concentration 50 mL L )

contained 12.5 mg of total organic matter (i.e., the final concentration  was  12.5  mg  L)1).  This  fraction  was  ac- counted by 5.0 mg of lipids (40% of total organic matter content), 1.5 mg of proteins (12.1%), 0.37 mg of organic phosphorus (3.2%), and 72.2 mg of carbohydrates (0.6%). 50 mL of solar oil contained 6.25 mg of lipids, 39.1 mg of proteins, and 1.1 mg of carbohydrates. The ratio of total organic C to total N and total P was 37:2:1 (w:w) in sun- screen, while in solar oil the C:N ratio was 100:1.



Long-Term Experiment


Viral And Bacterial Abundance. During the entire study period viral abundance in treated microcosms was sig- nificantly higher than control values (ANOVA control versus each different treatment, p < 0.05, except in ‘‘Oil’’ microcosms; Fig. 1). Viral abundance in ‘‘Sunscreen’’ (on average 16.6 · 109 viruses L)1) and ‘‘Sunscreen max’’ (on average  42.1  ·  109  viruses  L)1)  was  20–40  times  higher than  in  the  control  (on  average,  2.7  ·  109  viruses  L)1). Viral abundance in ‘‘Oil’’ (4.6 · 109 viruses L)1) and ‘‘Oil max’’ (10.2 · 109 viruses L)1) exceeded values observed in


2 d), whereas highest values in ‘‘Sunscreen’’ and ‘‘Sun- screen max’’ were observed after 15 and 31 days (2.6 and

95.8 · 109 viruses L)1, respectively).

While the input of 50 mL L)1 of sunscreen did not cause any significant change in bacterial abundance (compared with the control), the addition of 500 mL L)1 of sunscreen caused a significant increase (59 versus 35 · 108 cells L)1 in the control; ANOVA, p < 0.01).

No significant changes in bacterial abundance were observed  following  the  addition  of  50  and  500  mL  L)1  of solar  oil  (27  and  371  · 108  cells  L)1,  respectively).  In  all microcosms, bacterial abundance reached a  peak  within  the first 2 days of the experiments, except for ‘‘Sunscreen max’’ (8.5 · 109 cells L)1), in which a peak was reached at day 5 (Table 1).

Values of bacterial cell biomass are reported in Table 2. No significant changes in bacterial cell biomass were ob- served among different microcosms (ranging, on average, from 76.1 to 79.8 fgC cell)1) except in Sunscreen, in which bacterial  cell  biomass  (on  average,  88.3  fgC  cell)1)  was significantly higher than in the control (ANOVA, p < 0.05). The ratio of viruses to bacteria (VBR) was significantly enhanced after treatment with sunscreen and solar oil (on


Table 1. Total bacterial number, enzymatic activities, and bacterial C production



Bacterial number


Bacterial biomass








Bacterial C prod.




(108 cells L)1)

(lgC L)1)

(nmol L)1h)1)

(nmol L)1h)1)

(nmol L)1h)1)

(mgC L)1h)1)



30.4 ± 4.6

252.9 ± 34.3

47.1 ± 7.2

429.6 ± 28.4

420.4 ± 53.7

2.0 ± 0.2



2.8 – 69.3

24.7 – 747.5

5.0 – 81.1

152.2 – 1033.5

265.1 – 670.5

1.2 – 3.2



42.7 ± 3.7

321.2 ± 30.3

54.6 ± 5.2

1298.9 ± 153.3

698.0 ± 126.9

2.6 ± 0.3



4.5 – 98.2

52.4 – 553.0

12.8 – 88.9

217.5 – 2674.9

374.1 – 1061.8

0.6 – 5.9



27.2 ± 4.7

192.1 ± 33.3

53.9 ± 3.5

700.4 ± 61.9

576.5 ± 40.3

1.6 ± 0.1



1.2 – 67.7

11.3 – 525.3

11.5 – 90.4

297.6 – 1253.3

388.0 – 826.9

0.6 – 2.3



58.8 ± 9.3

461.6 ± 71.1

54.5 ± 3.6

1783.6 ± 130.3

2193.8 ± 143.1

7.5 ± nd



25.4 – 85.2

249.9 – 828.3

7.2 – 97.1

240.6 – 3049.0

114.9 – 4725.6

1.7 – 18.3

Oil max


37.1 ± 4.7

277.5 ± 33.3

63.4 ± 13.6

1268.9 ± 70.2

737.7 ± 43.4

3.6 ± nd



24.5 – 67.6

155.6 – 440.8

15.5 – 109.9

659.2 – 2747.5

459.2 – 1477.5

1.2 – 12.0



average, ANOVA, p < 0.01). The VBR ranged from 0.2 to

7.5 in the control (on average, 1.6), from 1.8 to 41.6 in the ‘‘Sunscreen’’ bag (on average, 8.9), from 1.4 to 26 in the ‘‘Sunscreen max’’ bag (on average, 8.5), and from 0.9 to

12.7 and from 1.7 to 5.3 in ‘‘Oil’’ and ‘‘Oil max’’ micro- cosms (on average, 4.0 and 3.2, respectively; Fig. 2).


Enzymatic activity. b-D-Glucosidase  activity  (Table  1)  was significantly higher in all microcosms with sun products added than in the control (ANOVA, all p < 0.05). Also aminopeptidase (ranging, on average, from 429.0 to 1808.8 nmol L)1 h)1 in the control and ‘‘Sunscreen max,’’ respectively) and phosphatase (ranging, on average, from

420.4 to 2193.8 nmol L)1 h)1 in the control and ‘‘Sunscreen max,’’ respectively) activities increased significantly in all microcosms after the addition of sunscreen products (ANOVA, p < 0.001).

Enzymatic activities normalized per cell are reported in Table 2. b-D-Glucosidase activity per cell was generally higher in the control (8.6 · 10)9 nmol cell)1 h)1) than after addition of sunscreen products, and was significantly re- duced  in  ‘‘Sunscreen  max’’  (4.1  ·  10)9  nmol  cell)1  h)1, ANOVA,  p  <  0.05)  and  ‘‘Oil’’  microcosms  (0.23  ·  10)9


nmol  cell)1  h)1,  ANOVA,  p  <  0.05),  whereas  in  ‘‘Sunsc- reen’’ (8.8 · 10)9 nmol cell)1 h)1) and ‘‘Oil max’’ micro- cosms   (5.5   ·  10)9   nmol   cell)1   h)1)   values   were   not significantly different from the control. Conversely, ami-

nopeptidase activity per bacterium was higher than in the control after the addition of sun product (2.2, 3.2, 3.4, 4.0, and  4.0  ·  10)7  nmol  cell)1  h)1  in  control,  ‘‘Sunscreen,’’ ‘‘Sunscreen max,’’ ‘‘Oil,’’ and ‘‘Oil max,’’ respectively; all ANOVA, p < 0.001, except ‘‘Sunscreen,’’ where ANOVA yielded p < 0.05). Finally, the addition of moderate con- centrations of sun products into microcosms did not change phosphatase activity per cell (range, 3.1–3.4 · 10)7 nmol  cell)1  h)1),  but  a  strong  reduction  of  phosphatase activity per cell was observed in ‘‘Sunscreen max’’ (0.1 · 10)7 nmol cell)1 h)1, ANOVA, p < 0.05).


Bacterial Production. When compared with the control (1.9 mgC L)1 h)1), bacterial C production was significantly enhanced  in  both  ‘‘Sunscreen  max’’  (8.0  mgC  L)1  h)1, ANOVA,  p  <  0.001)  and  ‘‘Oil  max’’  (3.5  mgC  L)1  h)1,

ANOVA, p < 0.05) microcosms, but was also found to be inhibited   in   the   ‘‘Oil’’   microcosm   (1.5   mgC   L)1   h)1,



Table 2. Bacterial cell biomass, enzymatic activities, and carbon production per cell (Std = standard deviation)




(fgC cell)1)

cell (nmol cell)1 h)1)

cell (nmol cell)1 h)1)

(nmol cell)1 h)1)

cell (fgC cell)1 h)1)

Control            Average

76.1 ± nd

8.60E-09 ± 3.99E-09

2.25E-07 ± 4.74E-08

3.42E-07 ± 5.76E-08

1.6 ± 0.3


66.1 – 88.7

1.58E-10 – 3.52E-08

2.06E-08 – 5.65E-07

3.93E-08 – 1.61E-06

0.1 – 4.8

Sunscreen       Average

88.3 ± nd

8.80E-09 ± 2.35E-09

3.22E-07 ± 3.95E-08

3.41E-07 ± 2.43E-08

0.8 ± 0.5


56.3 – 129.3

1.52E-09 – 3.51E-08

1.13E-07 – 6.47E-07

4.92E-08 – 1.24E-06

0.1 – 2.6

Oil                    Average

78.0 ± nd

4.10E-09 ± 9.34E-10

3.37E-07 ± 7.20E-08

4.33E-07 ± 9.80E-08

1.4 ± 0.4


59.9 – 102.1

8.73E-10 – 8.68E-09

4.52E-08 – 7.24E-07

2.19E-08 – 1.10E-06

0.0 – 4.2

SSmax             Average

79.8 ± nd

2.28E-08 ± 8.53E-09

6.96E-07 ± 1.27E-07

1.08E-06 ± 1.75E-07

1.2 ± 0.5


62.9 – 112.7

1.88E-09 – 1.15E-07

1.80E-07 – 2.88E-06

9.38E-08 – 6.41E-06

0.2 – 4.7

Oil max           Average

76.7 ± nd

5.47E-09 ± 3.17E-09

4.08E-07 ± 6.34E-08

2.11E-07 ± 3.24E-08

1.0 ± 0.1


61.7 – 98.4

2.43E-09 – 1.27E-08

1.95E-07 – 8.73E-07

6.35E-08 – 4.40E-07

0.2 – 3.8



Bacterial cell biomass


Glucosidase per


Aminopeptidase per


Phosphatase per cell


C production per




















Fig. 2. Long-term changes of VBR in the contror (Ctrl), ‘‘Sunscreen’’ (SS), and ‘‘Sunscreen max’’ (SSmax) microcosms (a), and VBR changes in the control (Ctrl), ‘‘Oil’’ (Oil), and ‘‘Oil max’’ (Oilmax) microcosms (b). Standard deviations (each value represents 3 replicate analyses of 3 microcosms) are reported at all sampling times (not visible when bars are within the mark).



ANOVA, p < 0.01). Bacterial C production per cell was significantly lower than in the control (1.6 fgC cell)1 h)1) only  in  the  ‘‘Sunscreen’’  microcosm  (0.8  fgC  cell)1  h)1, ANOVA, p < 0.05), but did not change significantly in ‘‘Sunscreen max’’ (1.4 fgC cell)1 h)1), ‘‘Oil’’ (1.2 fgC cell)1 h)1), and ‘‘Oil max’’ (1.0 fgC cell)1 h)1) microcosms.


Short-Term Experiment


Viral and Bacterial Abundance. Viral and bacterial abundance in the control did not change significantly with time (i.e., 120 min of incubation; Fig. 3a). In ‘‘Sunscreen max’’  viral  abundance  (on  average,  26.2  · 109  virus  L)1) was about one order of magnitude higher than in the control (ANOVA, p < 0.001), while bacterial abundance slightly decreased (on average, 2.1 · 109 cells L)1; ANOVA, p < 0.001; Fig. 3b). In comparison with the control, also in the ‘‘Oil max’’ treatment viral abundance (on average, 5.6

  • 109 viruses  L)1)  increased  (ANOVA,  p  <  001),  while

bacterial  abundance  (on  average,  1.9  · 109  cells  L)1)  de- creased (ANOVA, p < 0.001). In the ‘‘Sunscreen max,’’ viral and bacterial abundance displayed a synchronous oscillation with a period of 30, 18, 18, 24, 12, 18 min (on


average, 20 min). The ratio of viruses to bacteria increased significantly from the control (range, 0.7–1.3; on average 0.9) to the ‘‘Oil max’’ (range, 2.0–3.8; on average 2.9) to the ‘‘Sunscreen max’’ (range, 6.5–10.4; on average 7.8).




Effects of Sunscreen Products on Microbial Loop Structure


Several components of sunscreen products have been proved to be toxic and harmful for aquatic organisms, and active ingredients, being lipophilic and resistant to deg- radation in aquatic environments, are known to bio-ac- cumulate in different fish species [18]. We report here that sunscreen products inoculated into microcosms, even at low concentrations, affected microbial community struc- ture. Since the first days of incubation, viral abundance in all microcosms with sun products (except in the ‘‘Oil’’ microcosms) displayed significantly higher values than in the control (ANOVA, p < 0.001). However, at equal con- centrations of solar oil and sunscreen, the impact on viral abundance was 5–10 times higher in microcosms treated with sunscreen than in those treated with solar oil. The






























Fig. 3.   Short-term variations of bacterial abundance in control, ‘‘Oil max’’ and ‘‘Sunscreen max’’ microcosms (a) and viral abundance   in control, ‘‘Oil max,’’ and ‘‘Sunscreen max’’ microcosms (b). Standard deviations (each value represents 3 replicate analyses of 3 microcosms) are reported at all sampling times (not visible when bars are within the mark).




effect of cosmetic sun products was even more  evident after 10 days. Viral abundance remained significantly higher than in the control for 3 months.

The two cosmetic sun products had a different impact on bacterioplankton. In fact, bacterial abundance slightly decreased or remained unvaried after solar oil input, at both low and high concentrations, while it increased sig- nificantly (by ~70%) after sunscreen addition at high concentrations (ANOVA, p < 0.001). Finally, bacterial cell biomass was generally unaffected by addition of sunscreen products, except in the ‘‘Sunscreen’’ microcosm, where it increased significantly compared with the control.

The design of the short-term experiment was defined in order to clarify whether observed changes between treated and untreated microcosms were due to an erratic short- term fluctuation in viral abundance and to quantify the fraction of the bacterial community with lysogenic infec- tion. Our short-term experiments in replicate microcosms started after 15 days of incubation of the collected sea- water samples (i.e., when the effects of the sunscreen in- oculation were evident). Also previous studies adopted


this strategy to avoid the possibility of experimental arti- facts [8]. In ‘‘Sunscreen max’’ microcosms the increased viral abundance displayed a cyclic fluctuation, which was much less evident in the control and ‘‘Oil max’’ micro- cosms. The observed rates of viral increase in ‘‘Sunscreen max’’ are 1–2 orders of magnitude higher than any other rates of viral production and removal observed in natural marine environments, confirming experimental results carried out on Norwegian temperate seawater [8], and suggesting that sunscreen inoculation induced a lytic cycle in lysogenic bacteria [38]. We calculated the theoretical frequency (n) for small amplitudes in viral abundance assuming the host–phage bag as a predator–prey bag with

constant prey growth rate m and constant predator death rate d, as follows: n = 2pmd. With a bacterial growth rate (0.42 d)1 as calculated in this study) and assuming a viral ‘‘death’’ rate of 100 d)1 [8, 24], we obtain a frequency of

1.70  h)1,  whereas  the  experimental  frequency  of  oscilla- tions  was  3.0  h)1.  This  discrepancy  indicates  that  in  our microcosms viral dead rates could be approximately half or the one assumed. Moreover, since estimates of bacterial


growth rates depend upon the conversion factor used to estimate their biomass, actual viral dead rates could be even lower than those estimated in the present study.

In a diverse bacterial community with many host– phage microcosms, one would not necessarily be able to detect oscillations in total viral and bacterial numbers [7], unless virus production in different host–phage system is synchronized [4]. Although we cannot rule out other possibilities (i.e., systems dominated by a few different host phages or phages with a broad host range, [8]), it is possible that oscillations in viral abundance are a conse- quence of lysogenic bacterial infection. Since in our ex- periment we observed that the frequency of oscillation of the viral number increased from the control (<1 h)1) to the ‘‘Oil max’’ (~1 h)1), to the ‘‘Sunscreen max’’ (3.0 h)1) microcosms, and that bacterial and viral abundance dis- played synchronous oscillations, we can hypothesize that such dynamics of viral abundance are due to synchronous lysis and release of viral particles from lysogenic bacteria. At the same time, the significant increase of viral abundance after sunscreen treatment indicates the pres- ence of substances inducing prophages. Among mutagenic agents inducing lytic replication in lysogenic bacteria, the effect of mitomycin C are well known; however, mitomy- cin C is not naturally found in marine environments [1]. Other studies demonstrated that also aromatic hydrocar- bons, PCB mixtures, the pesticide Aroclor 1248, and pes- ticide mixtures cause induction of natural prophage populations [11, 26]. These studies, assuming an average burst size of 30, estimated that lysogenic bacteria ac- counted for 1.3% (in seawater samples treated with Aroclor 1248) to 120.8% (pesticide mixture) of natural bacterial assemblages. Since prophage induction may oc- cur even when there are only small changes in bacterial abundance [11], according to other studies, we based our estimates of inducible lysogenic bacteria on significant

increases in viral abundance alone [11, 26].

In order to estimate the fraction of lysogenic bacteria,  we assumed a burst size of 30 (repeatedly utilized in previous studies [11, 26]), and we found that 13–24% of total bacterial abundance was induced to a lytic cycle after sunscreen addition, whereas 6–9% of bacterial abundance was induced to a lytic cycle after solar oil addition. The burst size utilized for these calculations agrees well with  the burst size we estimated during the short-term exper- iment on the basis of bacterial C production. To do this   first we calculated bacterial turnover rates (as bacterial C production divided by bacterial biomass). Then we esti-


mated bacterial mortality as the difference between in- crease of bacterial abundance expected from bacterial turnover and abundance actually observed in the experi- ment. Finally, we estimated the burst size as the ratio between increased viral abundance and bacterial mortali- ty. In this way, we estimated a burst size of 119.5 and 29.1 after addition of sunscreen and oil at high concentrations, respectively. These results fall perfectly within the range of burst size values reported for oligotrophic and meso- trophic waters [9, 24, 36–38].

The effect of prophagic induction was evident at both low and high sunscreen concentrations, indicating that sunscreen release in coastal waters might have a signifi- cant impact even at very low levels. Values reported in this study are similar to those obtained after the addition of mytomicin C and mixture of PCB and pesticide in sea- water samples collected in different environments [11], indicating that sun products might significantly induce the lytic cycle in lysogenic bacteria.

Microbial response to the addition of cosmetic sun products can be due either to the presence of micropol- lutants (retinoids, imidazolidinyl urea, EDTA, aromatic hydrocarbons, benzoic acid, and PABA) or to the input of organic matter contained in the sun product. However, we can exclude that the microbial response was solely due to the organic matter content of solar products, since organic matter concentrations inoculated fall within the range naturally observed in coastal seawater samples [14]. Lipids represented, quantitatively, the most important compo- nent added to all experimental microcosms. Lipids are a class of compounds whose biological availability varies depending on their characteristics. However, viral increase and changes in VBR were far less evident in solar-oil treatments than in sunscreen treatments, despite the much higher lipid concentrations of the former. Therefore, ob- served changes in viral abundance and VBR are unlikely to be driven by the organic component and should be at- tributed to sun-blocking components added to the care products.


Biogeochemical Implications of Sunscreen Impact on Microbial Loop Functioning

Enzymatic activities are essential for the breakdown and subsequent uptake of organic matter by bacteria. There- fore, changes in enzymatic activities  have  consequences for biogeochemical cycling of organic matter, and there is indirect evidence that enzymatic activities might affect


viral decay [30]. All enzymatic activities determined in this study (b-glucosidase, aminopeptidase, and alkaline phos- phatase) increased after sunscreen and solar oil treatment. The enhancement of enzymatic activity varied according to concentration and typology of the sun product and was generally more evident for aminopeptidase and phospha- tase activities. Other insights into the impact of sunscreen products on microbial loop functioning have been gath- ered by calculating the specific bacterial enzymatic activ- ity. The significant increase of the per-cell aminopeptidase activity after sunscreen and solar oil supplementing at high concentrations, versus a decreased per-cell b-gluco- sidase activity, indicate that the addition of sun products affected selectively enzymatic activities. Therefore, the introduction of sun products in seawater might have also important ecological implications enhancing N and P cy- cling and reducing C mobilization. This fact, together with the enhanced viral development induced by solar prod- ucts, provides new bases for understanding virus–bacterial interactions in anthropogenically impacted systems [22].





We thank E. Manini and M. Armeni for laboratory assis- tance. The critical reading and suggestions of two anony- mous reviewers, A. Dell’Anno and A. Pusceddu contributed to improve the quality and readability of this article. This research was partially supported by grants of the Ministry for Environment, within the frame of the Program MAT (Mucilage in the Adriatic and Tyrrhenian Sea).





  1. Ackermann HW, DoBow MS (1987) Viruses of Prokaryotes: Natural Groups of Bacteriophages. CRC Press, Boca Raton, FL
  2. Amourex I, Pesando D, Noel H, Girard JP (1999) Mechanism of cytotoxicity by cosmetic ingredients in sea urchin Arch Environ Contam Toxicol 36:28–37
  3. Aspila KI, Agemiam H, Chau ASY (1976) A semi-automated method for the determination of inorganic,  organic,  and total phosphate in Analyst 101:187–192
  4. Bergh ØKY, Borsheim G, Bratbak G, Heldal M (1989) High abundance of viruses found in aquatic Nature 340:467–468
  5. Bettarel Y, Sime-Ngando T, Amblard C, Laveran H (2000) A comparison of methods for counting viruses in aquatic Appl Environ Microbiol 66:2283–2289


  1. Bligh EG, Dyer W (1959) A rapid method for total lipid extraction and purification. Can J Biochem Physiol 37:911– 917
  2. Borgmann U, Norwood WP (1995) EDTA toxicity and background concentrations of copper and zinc in Hyalella azteca. Can J Fish Aquat Sci 52:875–881
  3. Bratbak G, Heldal M, Thingstad TF, Tuomi P (1996) Dy- namics of viruses abundance in coastal seawater. FEMS Microbiol Ecol 19:263–269
  4. Bratbak G, Heldal M, Thingstad TF, Riemann B, Haslund OH (1992) Incorporation of viruses into the budget of mi- crobial C-transfer. A first approach. Mar Ecol Prog Ser 83:273–280
  5. Chrost RJ (1991) Measurement of enzyme kinetics in water samples: effect of freezing and soluble stabilizer. Mar Ecol Prog Ser 70:93–100
  6. Cochran PK, Kellog H, Paul J (1998) Prophage induction of indigenous marine lysogenic bacteria by environmental pollutants. Mar Ecol Prog Ser 164:125–133
  7. Corinaldesi C (2001) Sunscreens and marine pollution: ex- perimental study on sunscreen product effects on coastal marine Thesis, University of Ancona, Ancona, Italy
  8. Danovaro R, Armeni M, Dell’Anno A, Fabiano M, Manini E, Marrale D, Pusceddu A, Vanucci S (2001) Small scale dis- tribution of bacteria, enzymatic activity and organic matter in coastal sediments. Microb Ecol 42:177–185
  9. Danovaro R, Dell’Anno A, Pusceddu A, Marrale D, Della Croce N, Fabiano M, Tselepides A (2000) Biochemical composition of pico-, nano- and micro-particulate organic matter and bacterioplankton biomass in the oligotrophic Cretan Sea (NE Mediterranean). Progr Oceanogr 46:279–310
  10. Danovaro R, Dell’Anno A, Trucco A, Serresi M, Vanucci S (2001) Determination of virus abundance in marine sedi- Appl Environ Microbiol 67:384–387
  11. Danovaro R, Marrale D, Della Croce N, Dell’Anno A, Fabi- ano M (1998) Heterotrophic nanoflagellates, bacteria and labile organic compounds in continental shelf and deep-sea sediments of the Eastern Mediterranean. Microb Ecol 35:244–255
  12. Danovaro R, Mei ML, Dell’Anno A (2002) Higher abundance of bacteria than viruses in deep Mediterranean sediments. Appl Environ Microb 68:1468–1472
  13. Daughton CG, Ternes TA (1999) Pharmaceuticals and per- sonal care products in the environment: agents of subtle change? Environ Health Perspect 17:907–944
  14. Dubois M, Giles K, Hamilton JK, Reberts PA, Smith F (1956) Colorimetric method for determination of sugars and related substances. Anal Chem 28:350–356
  15. Fry JA (1990) Direct methods and biomass estimation. Methods in Microbiology, Vol 22, Academic Press, New York, pp 41–85
  16. Fuhrman JA, Noble RT (1999) Causative agents of bacterial mortality and the consequences to marine food webs. In: CR Bell, M Brylinsky, P Johnson Green (eds) Microbial Bio-


systems: New Frontiers. Proc VIII Int Symp Microb Ecol. Halifax, Canada

  1. Fuhrman JA (1999) Marine virus and their biogeochemical and ecological effects. Nature 399:541–548
  2. Hartree EF (1972) Determination of proteins: modification of the Lowry methods that give a linear photometric re- Anal Biochem 48:422–427
  3. Heldal M, Bratbak G (1991) Production and decay of viruses in aquatic environments. Mar Ecol Prog Ser 72:205–212
  4. Hoppe HG (1993) Use of fluorogenic model substrates for extracellular enzyme activity (EEA) of bacteria. In: PF Kemp, EB Sherr, JJ , Cole (eds) Handbook of Methods in Aquatic Ecology. Lewis, Boca Raton, FL, pp 423–431
  5. Jang SC, Paul JH (1996) Occurrence of lysogenic bacteria in marine microbial community as determined by prophage induction. Mar Ecol Prog Ser 142:27–38
  6. Jang SC, Paul JH (1994) Seasonal and diel abundance of viruses and occurrence of lysogeny/bacteriocinogeny in the marine environment. Mar Ecol Prog Ser 104:163–172
  7. Maden M (1996) Retinoic acid in development and regen- eration. J Biosciences 2:299–312
  8. Marsh JB, Weinstein WJ (1966) A simple charring method for determination of J Lipids Res 7:574–576
  9. Noble RT, Fuhrman JA (1997) Virus decay and its causes in coastal waters. Appl Environ Microbiol 63:77–83
  10. Noble RT, Fuhrman JA (1998) Use of SYBR Green I for rapid epifluorescence counts of marine viruses and bacteria. Aqua Microb Ecol 14:113–118
  11. Parker JC (1983) A comparison of methods used for mea- surement of organic matter in Chem Ecol 1:201– 210


  1. Routledge EJ, Parker J, Odum J, Ashby J, Sumpter JP (1998) Some alkyl hydroxy benzoate preservatives (parabens) are Toxicol Appl Pharmacol 153:12–19
  2. Simon, M, Azam F (1989) Protein content and protein synthesis rates of planktonic marine Mar Ecol Prog Ser 51:201–203
  3. Strickland JD, Parsons TR (1968) A practical handbook of seawater Bull Fishery Board Can 167:1–312
  4. Weinbauer MG, Fuks D, Peduzzi P (1993) Distribution of viruses and dissolved DNA along a coastal trophic gradient in the northern Adriatic Sea. Appl Environ Microbiol 59:4074–4082
  5. Weinbauer MG, Peduzzi P (1994) Frequency size and dis- tribution of bacteriophages in different marine bacterial morphotypes. Mar Ecol Prog Ser 108:11–20
  6. Wilcox RM, Fuhrman JA (1994) Bacterial viruses in coastal seawater: lytic rather than lysogenic production. Mar Ecol Prog Ser 114:35–45
  7. Wilhelm SW, Weinbauer MG, Suttle CA, Jeffrey WH (1998) The role of sunlight in the removal and repair of viruses in the sea. Limnol Oceanogr 43:586
  8. Wilhelm SW, Suttle CA (2000) Viruses as regulators of nu- trient cycles in aquatic environments. In: CR Bell, M Bry- linsky, P Johnson Green (eds) Microbial Biosystems: New Proc VIII Int Symp Microb Ecol, Halifax, Canada, pp 551–556
  9. Wommack KE, Colwell RR (2000) Virioplankton: viruses in aquatic Microbiol Mol Biol Ann Rev 64:69–114
  10. Wommack KE, Hill RT, Kessel M, Russek-Cohen E, Colwell RR (1992) Distribution of viruses in the Chesapeake Appl Environ Microbiol 58:2965–2970

About the author

Dr. Howard Dryden

Dr. Howard Dryden

Dr. Dryden has unique knowledge combination of biology, chemistry and technology and is the inventor of the activated, bio-resistant filter media AFM®. Dr. Dryden is one of the world`s leading experts in sustainable water treatment.

comments powered by Disqus

Wherever you are and wherever you live, there is no safe haven from the toxic wave of chemical pollution

Dr. Howard Dryden, CSO

Goes Foundation

3/2 Boroughloch Square

Clean Water Wave

The Meadows