Inhibition of lentivirus replication by aqueous extracts of Prunella vulgaris
© Brindley et al. 2009
Received: 21 November 2008
Accepted: 20 January 2009
Published: 20 January 2009
Skip to main content
© Brindley et al. 2009
Received: 21 November 2008
Accepted: 20 January 2009
Published: 20 January 2009
Various members of the mint family have been used historically in Chinese and Native American medicine. Many of these same family members, including Prunella vulgaris, have been reported to have anti-viral activities. To further characterize the anti-lentiviral activities of P. vulgaris, water and ethanol extractions were tested for their ability to inhibit equine infectious anemia virus (EIAV) replication.
Aqueous extracts contained more anti-viral activity than did ethanol extracts, displaying potent anti-lentiviral activity against virus in cell lines as well as in primary cell cultures with little to no cellular cytotoxicity. Time-of-addition studies demonstrated that the extracts were effective when added during the first four h of the viral life cycle, suggesting that the botanical constituents were targeting the virion itself or early entry events. Further analysis revealed that the extracts did not destroy EIAV virion integrity, but prevented viral particles from binding to the surface of permissive cells. Modest levels of anti-EIAV activity were also detected when the cells were treated with the extracts prior to infection, indicating that anti-EIAV botanical constituents could interact with both viral particles and permissive cells to interfere with infectivity. Size fractionation of the extract demonstrated that eight of the nine fractions generated from aqueous extracts displayed anti-viral activity. Separation of ethanol soluble and insoluble compounds in the eight active fractions revealed that ethanol-soluble constituents were responsible for the anti-viral activity in one fraction whereas ethanol-insoluble constituents were important for the anti-viral activity in two of the other fractions. In three of the five fractions that lost activity upon sub-fractionation, anti-viral activity was restored upon reconstitution of the fractions, indicating that synergistic anti-viral activity is present in several of the fractions.
Our findings indicate that multiple Prunella constituents have profound anti-viral activity against EIAV, providing additional evidence of the broad anti-viral abilities of these extracts. The ability of the aqueous extracts to prevent entry of viral particles into permissive cells suggests that these extracts may function as promising microbicides against lentiviruses.
P. vulgaris, commonly known as "self-heal", is a low-growing perennial herb with worldwide distribution. The herb is a member of the mint family Lamiaceae. Salves, teas, and extracts made from the plant have been used to treat wounds, inflammation, and other minor body disorders by both the Chinese and Native Americans [1, 2].
Various bioactive constituents have been identified in extracts of P. vulgaris. These include phenolic constituents, complex carbohydrates and more hydrophobic metabolites such as triterpenes. The abundant polysaccharides present in P. vulgaris are readily extracted by water and have a number of reported biological activities [3, 4], and several of the triterpenes have been identified with significant anti-inflammatory activity . Large quantities of anti-oxidants are known to be present in aqueous Prunella extracts with the polyphenolic compound, rosmarinic acid, being one of the most abundant of these constituents [6, 7]. Rosmarinic acid has also been shown to have anti-inflammatory activity as a result of specific inhibition of T cell signaling and an impact on glucose metabolism [8–10].
Prunella extracts have been reported to contain anti-viral and anti-bacterial properties, although constituents responsible for these activities are incompletely characterized to date [7, 11, 12]. Recent research has confirmed that anionic polysaccharides in aqueous extracts of P. vulgaris can decrease the replication of herpes simplex virus-1 and -2 (HSV-1, HSV-2) by preventing viral binding to cells [11, 13–15]. P. vulgaris extracts have also been shown to contain anti-HIV activity. Studies have identified inhibition of HIV infection at steps of virus binding , fusion , reverse transcription , integration , and protease function . Many of these studies identified Prunella antiviral activity through high through-put screens for specific viral protein targets in in vitro assays. While constituents in Prunella may be effective against these numerous anti-HIV targets in vitro, inhibition of the specific targets responsible for anti-HIV activity of Prunella in cells remains unclear. Identification of constituents of P. vulgaris that confer the inhibition to HIV-1 is limited to the water soluble, 10 kDa polysaccharide, Prunellin, that interferes with HIV-1 virion binding to permissive cells [16, 20]. Rosmarinic acid extracted from other botanicals has proved effective against HIV-1 integrase , but the role of this polyphenol in the anti-retroviral activities of Prunella extracts has not been explored. Additional members of the Lamiaceae, such as peppermint and lemon balm, are also known to have anti-viral activities, but specific constituents responsible for those activities remain unidentified [13, 22].
In this study we sought to examine the breadth of the anti-lentiviral activity of water and ethanol extracts from several P. vulgaris accessions by investigating their ability to inhibit replication of equine infectious anemia virus (EIAV). Water extracts of two of the accessions that had the greatest anti-viral activity were determined to interfere with virus binding and uptake. Our studies identified several different constituents present in the aqueous extracts that had significant activity against EIAV. Our findings suggest that this extract may serve as an effective microbicide against lentiviruses.
All Prunella vulgaris plant samples were provided by the North Central Regional Plant Introduction Station (NCRPIS, Ames, IA) of the Agricultural Research Service of the U.S Department of Agriculture. All samples utilized in experiments were produced from populations collected from North Carolina or Missouri in October, 2004 on a collection trip sponsored by the USDA/NCRPIS/ISU/NIH. Both seed and voucher specimens were collected from all original sites and specimens were keyed to species . Seeds from accessions Ames 27664, 27665, 27666 and 27748 were germinated in Petri plates at 25°C, transferred to flats in a greenhouse (20–25°C) before final field transfer into individual control pollinated screened cages in Ames, IA. Upper flowering portions of 14 month old plants were harvested, dried for 1 week at 38 °C in a forced-air dryer with constant humidity and ground (RTC-R301ULTRAB) for analysis. All voucher specimens representing both original and regenerated populations are stored in the Ada Hayden Herbarium, Iowa State University (Ames, IA: ISC). Seeds representing both original and regenerated populations are stored at the USDA NCRPIS under controlled conditions (-20°C, 4°C for regenerated samples). Information about the specific provenance of all accessions used for the experiments is available via the Germplasm Resources Information Network database at http://www.ars-grin.gov/npgs/acc/acc_queries.html.
All glassware was heated at 200°C for 2 h to destroy endotoxin.
One hundred mL of boiling, endotoxin-free water was poured over 6 g of dried P. vulgaris. The plant material was steeped, with stirring, for 1 h and filtered through a G6 glass fiber circle (Fisher Scientific) in a Buchner funnel. The filtrate was centrifuged at 10,000 × g for 20 minutes to remove any additional particulates. The extract was lyophilized, weighed, and re-dissolved in DMSO.
Six g of dried P. vulgaris was extracted with 500 mL of 95% ethanol via Soxhlet for 6 h. The extract was filtered, dried by rotary evaporation at < 40°C and then lyophilized. Extracts were resuspended in DMSO.
Two g of dry Prunella water extract, dissolved in 10 mL endotoxin-free water, was loaded onto a 2.5 × 75 cm Sephacryl 100HR column. Endotoxin-free water was used to elute the size-exclusion column. Two L of eluent was collected in 10 mL fractions collected for 72 h. Absorbance at 210 nm was measured for all fractions to monitor separation efficiency and identify peaks. Nine peaks were detected. Fractions composing these peaks were pooled and concentrated by lyophylization. Fractions were resuspended in endotoxin-free water.
All extracts and fractions were evaluated for endotoxin using the Chromogenic Limulus Amebocyte Lysate Test kit per manufacturer's instructions (Cambrex Bioscience Inc.). This assay is able to detect concentrations of endotoxin of 0.007 EU/mL or greater. All extracts had <0.007 EU/mL at the highest concentrations used in these studies. The fractions had slightly higher endotoxin levels; the highest amount of endotoxin present in the fractions when diluted for these studies was 0.023 EU/mL.
Sufficient 100% ethanol was added to each fraction to yield a 95% ethanol solution. These fractions were placed in a rotary shaker at room temperature for 1h. Fractions were centrifuged at 10,000 × g for 20 min. The ethanol-soluble supernatant was decanted, and the ethanol-insoluble pellet was redissolved in endotoxin-free water. Each sub-fraction was lyophilized and weighed and resuspended in endotoxin-free water.
Equine dermis cells (ED cells) (ATCC CCL57) were maintained in high glucose DMEM with 15% fetal calf serum (FCS). Primary equine umbilical vein endothelial cells (eUVEC) were also used in the EIAV studies and were maintained in high glucose DMEM with 40% FCS. All media were supplemented with penicillin and streptomycin.
Stocks of EIAV were generated in ED cells. Viral stocks of EIAVWSU5 , EIAVMA-1 , EIAVvMA-1c , EIAVSP19 , and EIAVTh1  from ED cell supernatants were harvested from cells that were >95% positive for EIAV antigen as determined by EIAV antigen immunostaining. Supernatants were centrifuged for 5 min at 13,500 × g to remove cell debris, aliquoted, and frozen at -80°C until needed. Viral titers were determined by infection of ED cells using the single round of infection assay described below.
Concentrations of Prunella stocks
Ames 27664 – water
Ames 27665 – water
Ames 27666 – water
Ames 27748 – water
Ames 27664 – ethanol
Ames 27665 – ethanol
Ames 27666 – ethanol
Ames 27748 – ethanol
Ethanol soluble 1
Ethanol soluble 2
Ethanol soluble 3
Ethanol soluble 4
Ethanol soluble 5
Ethanol soluble 6
Ethanol soluble 7
Ethanol soluble 8
Ethanol soluble 9
Ethanol insoluble 1
Ethanol insoluble 2
Ethanol insoluble 3
Ethanol insoluble 4
Ethanol insoluble 5
Ethanol insoluble 6
Ethanol insoluble 7
Ethanol insoluble 8
Ethanol insoluble 9
EIAVWSU5 was added to ED cells at an MOI of 0.005 in ED media. DMSO or extracts of P. vulgaris Ames 27664 or Ames 27748 extract was added to the well at 0, 1, 2, 3, 4, 6, and 8 h following infection at a final concentration of 0.2% DMSO (66 μg/mL of Ames 27664 or 62.4 μg/mL of Ames 27748). Forty h following infection the cells were fixed, immunostained for EIAV antigen and the EIAV positive cells enumerated.
EIAVWSU5 was bound to ED cells at 4°C for 1 h to permit binding, but prevent virion internalization. The cells were warmed to 37°C and DMSO or Prunella extract was added to the well at 0, 1, 2, 3, 4, 6, and 8 h following temperature shift at a final concentration of 0.2% DMSO (66 μg/mL of Ames 27664 or 62.4 μg/mL of Ames 27748). Forty h following infection the cells were fixed, immunostained for EIAV antigens and EIAV-positive cells enumerated.
EIAVWSU5 was bound to ED cells at 4°C for 1 h to permit binding, but prevent virion internalization. Unbound virus was removed, new media replaced, and the cells shifted to 37°C to promote internalization. At 0, 1, 2, 4, and 6 h following temperature shift, the cells were washed briefly in citrate acid buffer (pH 3.0) to inactivate any non-internalized virions. The citrate buffer was removed and cells were washed twice, and medium contain 0.2% of DMSO, 66 μg/mL of Ames 27664 extract, or 62.4 μg/mL of Ames 27748 extract was added to determine if the extracts had any inhibitory effect on virions that had already been internalized.
EIAV viral stock was incubated in DMEM with 10% fetal calf or DMEM 10% fetal calf plus 132 μg/mL of Ames 27664 extract or 126 μg/mL of Ames 27748 extract. The virus stock was maintained at 37°C and used to infect ED cells at various time points following extract exposure. The final concentration of Prunella when diluted on the cells was 0.44 μg/mL of Ames 27664 extract or 0.42 μg/mL of Ames 27748 extract. At 40 h following initiation of infection, the cells were fixed and immunostained for the production of EIAV proteins.
Virus was mixed with 132 μg/mL of Ames 27664 extract or 126 μg/mL of Ames 27748 extracts (final concentration of 0.4% DMSO) or fractions (100 ug/mL) and incubated with ED cells (MOI of 2) at 4°C for 2 h to permit binding, but prevent virion internalization. Unbound virions were removed and cells were washed with phosphate buffered saline (PBS) three times to ensure all unbound virions were removed from the cells. Each well was lysed in 50 μL of lysis buffer (50 mM Tris HCl (pH 8), 120 mM NaCl, and 0.5% NP40, and 1 U/mL of protease inhibitor cocktail (Sigma). The lysates were analyzed by immunoblotting for the presence of viral capsid to indicate virus binding as described below. Blots were re-probed for cellular β-tubulin to normalize for cellular input.
105 infectious particles of EIAVwsu5 were incubated at room temperature for 10 min with P. vulgaris Ames accession 27664 aqueous extract. Following the incubation, dilutions of the incubated virus were added to ED cells in a 48-well format and appropriate concentrations of extract were maintained on the cells for the duration of the experiment. Cells were fixed at 40 h following infection and immunostained as described above. Wells with serial dilutions containing between 10 and 250 virus positive cells were enumerated and back-calculations were made to obtain the numbers of infectious units of virus/mL.
Cell lysates were run on NuPAGE Novex Bis-Tris Mini Gels (Invitrogen) and transferred to nitrocellulose. EIAV capsid was detected using the 2085 sera (1:10000) and secondary anti-horse antisera (1:10000) that was used for immunostaining. Tubulin was detected by the E7 monoclonal antibody (1:2000) (NIH Developmental Studies Hybridoma Bank, University of Iowa) and sheep anti-mouse HRP secondary (GE Healthcare) (1:50,000). All immunoblots were visualized using WestDura (Pierce).
Sucrose step gradients were prepared by layering 250 μL aliquots of decreasing concentrations of sucrose (20%–60%) into 3 mL ultra centrifugation tubes. The gradients were allowed to equilibrate at 4°C for at least 3 h. Virions were treated with extracts (0.4%), 0.5% Triton-X 100 or DMSO for 1 h at 37°C and loaded onto the top of the gradients. Tubes were centrifuged for 16 h at 40,000 rpm in a SW60 rotor at 4°C and stopped without a brake. Two hundred and fifty microliter aliquots were collected beginning from the top of the tube and stored at -80°C until analyzed by immunoblotting.
Cells were plated and treated with extracts as described above. Forty h following treatment cell viability was monitored by ATPLite Assay (Packard Biosciences) per manufacturer's instructions.
Studies were performed at least three independent times except where noted in the figure legends. Means and standard errors of the mean are shown. Student's t-test was used to evaluate the statistical differences between treatments, utilizing the two-tailed distribution and two-sample equal-variance conditions. P-values were assessed by comparing the level of infectivity with treatment to the level of cytoxicity seen with that same treatment. P-values for viability were assessed by comparing the level of viability with treatment to the level of viability seen with DMSO or control. A significant difference was determined by a p-value of < 0.05 and significance was identified in each figure. If the p-value was > 0.05, the data were not considered statistically significantly different.
Aqueous extracts of Ames 27664 and 27748 were also tested for their ability to inhibit a variety of EIAV strains (Fig. 2B). Two tissue-culture adapted strains, EIAVMA1 and EIAVSP19, as well as a field isolate, EIAVTh1, and a variant superinfecting strain, EIAVvMA1c, were effectively inhibited by the extracts. Previous studies have demonstrated that EIAVvMA1c enters ED cells through an alternative pathway compared to its parental strain EIAVMA1 [28–30]. EIAVvMA1c enters ED cells through plasma membrane fusion whereas EIAVMA1 and other wild-type strains of EIAV enter ED cells through interaction with the cellular receptor ELR1 that is mediated by a low-pH dependent, clathrin-mediated endocytosis event. The observation that the P. vulgaris extracts inhibit both the wild-type strains and variant strain equivalently suggest that Prunella anti-viral activity is broadly inhibitory and does not block specific viral-entry events, such as viral glycoprotein/ELR1 interactions.
Percent of virus added that is sensitive to Prunella extracts
Time of addition (hr)
Pre bound virions
We also tested the ability of Prunella aqueous extracts to inhibit virions that have been internalized from the cell surface. Virions were bound to ED cells at 4°C, unbound virions were removed, fresh media replaced and the cells shifted to 37°C to promote virion internalization. At 0, 1, 2, 4, and 6 h following 37°C temperature shift, the cells were treated with citric acid buffer that inactivates all virions remaining on the cell surface. The cells were washed and media containing DMSO or extracts added to the cells and maintained for the 40 h infection. Internalized virions were not impacted by Prunella extracts (Fig. 4C). In total, our data suggest that the Prunella extracts inhibit EIAV infectivity by interfering with virus binding and subsequent requisite steps that occur prior to virion internalization. However, once the virions are internalized, the extract was not inhibitory.
To determine the reduction in particle infectivity by P. vulgaris extracts, we incubated 105 infectious virions of EIAVWSU5 with 25 to 100 μg/mL of extract for 10 min at room temperature. Virions were serially diluted in media containing the same concentration of extract and plated on ED cells. Virus infectivity was evaluated 40 h later. Incubation of virions with aqueous extract has a profound impact on virion infectivity with 100 μg/mL of extract resulting in greater than 3000-fold reduction in infectivity, indicating that the majority of the anti-viral effect seen is caused by the extracts interacting with the viral particles directly, rather than inhibiting later steps in the viral life-cycle (Fig. 5B).
To further characterize the anti-viral constituents in the fractions, ethanol precipitation of the nine fractions was performed to separate ethanol-soluble and insoluble compounds. Constituents present in the sub-fractions were weighed and resuspended in endotoxin-free water at concentrations that represented the same ratio between the soluble and insoluble constituents present in the original fraction (see Table 1 for concentrations). A single ethanol-soluble fraction, Fraction 6, showed significant inhibition of EIAV (Fig. 8B). Ethanol-insoluble sub-fractions 4 and 9 displayed potent anti-EIAV activity (Fig. 8C). Interesting, the ethanol precipitation of fractions 2, 3, 5, 7 and 8 resulted in complete loss of anti-EIAV activity in either sub-fraction.
To determine if the ethanol precipitation destroyed the anti-viral activity or if multiple constituents that were separated during sub-fractionation were required for activity, we performed reconstitution experiments. The soluble and insoluble sub-fractions were added together at concentrations found in the original fractions and tested for anti-viral activity (Fig. 8D). The anti-viral activity seen in the original fraction 2 and 3 was lost after sub-fractionation and was not reconstituted. Reconstitution of fractions 4 and 9 did not enhance the anti-viral activity over that observed with the ethanol insoluble sub-fraction alone. Fraction 6 from the ethanol-soluble sub-fraction displayed anti-viral activity; however, after reconstitution, anti-viral activity was enhanced. Surprisingly, anti-viral activity was restored in fractions 5, 7 and 8 after reconstitution, suggesting synergy between constituents is required for the anti-viral activity of these fractions.
This study identified anti-viral activity against the lentivirus EIAV in aqueous extracts of P. vulgaris. The primary mechanism of inhibition of viral replication targeted viral entry. The extracts dramatically reduced infectivity when incubated with the virions alone and interfered with the ability of virus to bind to permissive cells. However, entry of EIAV particles that were pre-bound to ED cells prior to exposure to the extract was also inhibited, suggesting the anti-viral activity was not limited to inhibition of viral binding, but also prevented additional external events that are required for subsequent internalization and/or fusion. These extracts were not blocking specific interactions between EIAV and permissive cells such as the interaction of the gp90 glycoprotein and the cellular receptor ELR1 since the extracts effectively blocked infection of EIAVvMA-1c which can utilize a different cellular receptor .
Our fractionation studies indicated that numerous Prunella constituents were present in the aqueous extracts that have inhibitory activity against EIAV. To begin to identify the individual constituents responsible for the anti-viral activity, we separated the aqueous extracts by size-exclusion chromatography and subsequently separated those fractions into ethanol-soluble and insoluble components. Initial fractionation of the extract by size was not highly informative since 8 of the 9 fractions retained anti-viral activity. With five of these fractions, separation of constituents by ethanol precipitation resulted in loss of all activity; whereas, the ethanol-soluble material from one fraction had anti-viral activity and ethanol-insoluble sub-fractions from two other fractions were active. The anti-viral activity found in fractions 4 and 9 were ethanol-insoluble suggesting that carbohydrates may be responsible for the activity. The activity found in fraction 6 was ethanol-soluble and therefore is likely to be polyphenolic in nature. Of the five fractions where activity was lost upon sub-fractionation, activity was reconstituted when the ethanol-soluble and insoluble sub-fractions were combined to regenerate fractions 5, 7 and 8. Our findings demonstrated that synergy between ethanol-soluble and insoluble constituents is necessary for the anti-viral activity in these fractions. Botanical constituents responsible for these anti-viral activities remain to be identified.
The extracts were found to effectively inhibit a range of EIAV strains from field isolates to a laboratory variant that can use a different cellular receptor to enter cells. The inhibition was observed in primary cells as well as a cell line. The breadth of the anti-viral activity of Prunella extracts was not unexpected since Prunella aqueous extract had previously been characterized to inhibit both the distantly related lentivirus HIV-1 and the unrelated DNA virus HSV-1 [11, 12, 16, 17]. One of these studies identified a 10 kDa sulfated carbohydrate Prunellin from Prunella extracts that inhibited HIV-1 entry [16, 20]. A carbohydrate of approximately that same size was responsible for inhibiting HSV-1 entry into cells . While not definitively demonstrated, it is likely that Prunellin is also responsible for the anti-HSV-1 activity. Prunellin may be responsible for anti-EIAV activity found in ethanol-insoluble fraction 4 or 9. However, the ability of ethanol-insoluble constituents present in two non-contiguous fractions to robustly inhibit infectivity implicates additional, currently unidentified carbohydrates in the anti-viral activity against EIAV.
Extracts from other Lamiaceae species have been shown to bind to HIV-1 particles interfering with HIV-1 entry into permissive cells . HIV-1 virions in the presence of extracts from lemon balm were shown to be denser in a sucrose gradient than virions in absence of extracts suggesting that extracts either altered the structure of the particle resulting in enhanced particle density or the constituents were bound to virions making the particles denser . In our study, EIAV virions were not destroyed by treatment with the extracts. Nor did we observe a change in EIAV virion density as had been reported for HIV-1/lemon balm extracts . While, it is likely that the Prunella extract binds to EIAV particles reducing productive, but non-specific interactions with target cells, this interaction did not significantly alter virion density. In addition, lemon balm extracts were not effective against HIV particles that were pre-bound to cells , a finding distinctly different from our observations that both binding events and post-binding events were affected by Prunella extracts.
The aqueous extracts from two of the Prunella accessions were significantly more inhibitory than extracts from two other accessions that were evaluated despite the fact that all four accessions were field grown in Iowa under similar conditions. The concentrations of the extracts used in this study could not account for this observation suggesting that there is extensive genotypic variation of Prunella in the field. The two extracts that were most effective were collected in areas of disturbed habitat adjacent to roads and are likely to have been recently introduced. In contrast, the extracts with weaker anti-EIAV activity were found in undisturbed, forested areas in North Carolina. Further genetic and metabolomic studies will be required to understand this potential constituent diversity within Prunella vulgaris.
equine infectious anemia virus
equine dermis cells
equine umbilical vein endothelial cells
human immunodeficiency virus-1
herpes simplex virus-1
equine lentivirus receptor-1
This publication was made possible by grant number 9P50AT004155-06 from the National Center for Complementary and Alternative Medicine (NCCAM) and Office of Dietary Supplements (ODS). Its contents are solely the responsibility of the authors and do not necessarily represent the official views of the NIEHS, NCCAM, or NIH. Mention of commercial brand names does not constitute an endorsement of any product by the U.S. Department of Agriculture or cooperating agencies.
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.