One hundred randomly selected cats of six months or older living in a cat sanctuary (Room 8 Memorial Cat Foundation (R8MCF), Riverside, CA, USA) (n = 62), in an adoption guarantee (formerly referred to as no-kill) shelter (Mary S. Roberts Pet Adoption Center (MSRPAC), Riverside, CA, USA) (n = 20), in homes (Riverside, CA, USA) (n = 12), and in an American Association of Feline Practitioners-certified cat friendly practice with gold status (Riverside Cat Hospital, Riverside, CA, USA) (n = 6) were included in this study. The R8MCF housed cats in five rooms, with on average 20–25 cats per room (averaging two meters squared by two and a half meters high per cat), whereas the MSRPAC housed cats in individual cages (averaging three meters squared by two meters high per cat). About 55% of the surface space at the R8CMF was outdoors, and the cats could freely move between indoors and outdoors. Cats at the MSRPAC were allowed outside of their cages into a 100 square meters open room to interact with each other daily for eight hours. All cats at both facilities had continuous access to hiding places, cat beds, shelves and cat trees, scratching boards, litter boxes and water. The cats at the R8CMF had continuous access to dry food and received wet food at least once daily, whereas the cats at the MSRPAC received dry food twice daily. At the R8CMF window fans were used to help cool the rooms, whereas at the MSRPAC a constant temperature of 24 °C was maintained.
Cats in isolation areas or with signs of upper respiratory disease were excluded. Data were missing for some cats because they were adopted while the study was in progress. Only cats for which we had data for all four stimuli (n = 95) were included in the response pattern analyses. Cat age, sex, and breed were recorded when available. Ages were available for 89 cats, although for seven senior cats it could only be reliably determined they were at least 10 years of age. To test whether responsiveness to the plant materials differed between younger, generally more playful, cats and older cats, the group of 89 cats was divided into a younger (≤4 years and 10 months, n = 45) and older (>4 years and 10 months, n = 44) group using the median age of these 89 cats as a cutoff between the groups. The average age for the younger group was 1 year and 11 months (range 7 months to 4 years and 10 months) and at least 9 years and 2 months (range 5 to 16 years) for the older group. Cats also were classified (assisted by staff and volunteers who helped socialize the cats) into three different behavioral categories: (i) scared or shy (defined as avoiding humans by hiding or not leaving their location after humans entered their room), (ii) intermediate (defined as showing an interest in human presence and enjoying being petted when approached by humans), and (iii) affectionate or friendly (defined as immediately approaching humans entering their room, wanting to be petted) to study the influence of social behavior on the response to the plant materials.
Dried, cut and sifted organic catnip (Nepeta cataria) leaves and flowers were obtained from Frontier (Norway, IA, USA) and Smarty Kat (San Rafael, CA, USA). Dried, cut and sifted organic valerian (Valeriana officinalis) root was purchased from Organic Bio Herbs (Woodland Park, NJ, USA) and Frontier (Norway, IA, USA). Tatarian honeysuckle (Lonicera tatarica) wood and sawdust were purchased from The Cat House Inc. (Calgary, AB, Canada). Powder from dried silver vine (Actinidia polygama) fruit galls was purchased from Smack (Nagoya, Japan) and Gendai Pharmaceutical (Tokyo, Japan). Dried normal silver vine fruit and silver vine fruit galls were kindly provided by Hangzhou Botanical Technology (Hangzhou, China). We were unable to obtain any Indian nettle material for this study. To prevent possible degradation of active compounds, all plant materials were ordered just prior to the start of the experiments. During the study period (early March till the middle of April in 2016) all plant materials were stored tightly sealed in their original bag, at room temperature and in the dark.
Cats were exposed to 0.5 to 1.0 g silver vine powder (corresponding to 1–2 packets as sold by the manufacturers) inside a thin, porous sock (sock hereinafter), or 0.5 to 1.0 g silver vine powder spread out on approximately 500 square cm of a piece of 0.5 square m of Frieze carpet (carpet hereinafter). This type of carpet was used because it allowed more direct exposure to the materials than did the socks, while still minimizing the ability of the cats to ingest the plant material. Equal volumes of catnip (5 g), valerian root (15 g) and Tatarian honeysuckle sawdust (15 g) were offered inside a sock. This amount of plant material filled the sock about half full, and allowed tying a knot to close off the sock. Tatarian honeysuckle also was offered as an approximately 10 cm long piece of wood (Fig. 2a). Catnip (5 g) and valerian root (15 g) also were offered on carpet. We chose these amounts and volumes of catnip, Tatarian honeysuckle and valerian root to ensure that the cats would be exposed to an amount of active compounds that would allow for a positive response. We did not use a similar volume of silver vine powder because this would greatly exceed the manufacturer’s recommendation. Furthermore, the high price and limited supply of commercially available silver vine powder outside of East Asia would prevent the use of such large quantities in everyday life (e.g., cat owners, shelters). Results from preliminary experiments demonstrated that 1 g of silver vine powder was enough to elicit a positive response in each of the cats tested (Additional file 1). Socks or carpet were placed on the floor of the cat’s environment, within one meter of the cat’s head, in its line of sight or in the center of the room, respectively. When a cat seemed unaware of the presence of the plant material, the sock was repositioned within sight of the cat, once. Plant materials were never forced onto any of the cats, nor were they repositioned when a cat walked away from it. Cats responding positively to the plant materials on the carpet also responded positively to the plant materials in the sock (personal observation), suggesting that the fabric did not significantly interfere with the exposure of the active compound(s) to the cats. All plant materials were stored inside separate, sealed plastic bags to prevent cross-contamination. Identical carpet without plant materials and an identical, empty sock were used as negative controls.
The majority of the cats were exposed to all four plant materials on three different days (2 to 16 days apart), at different times of the day (morning, afternoon or evening), for up to one hour. The minimum number of exposures for each cat was two, on different days, and the minimum exposure duration for each cat was ten minutes per exposure. All cats were exposed to the plant materials presented in a sock and the majority of them also were exposed to a piece of Tatarian honeysuckle wood, and once to the plant materials on the carpet. Each plant material was offered at separate times, in random order, with a washout period of at least five minutes in between to avoid recording false positive responses resulting from a carry-over effect. The length of this washout period was based on results from preliminary experiments (Additional file 1). New plant material was only offered to a cat when it would not or no longer show any of the behavior associated with the catnip response (see below) when offered a negative control sock. When a positive, complete response (described below) was recorded, the stimulus was removed from the cat. This was done to prevent possible response saturation (e.g., because an active compound may be present in more than one plant material or because the cat became fatigued) from negatively affecting responses to the plant materials tested subsequently. In addition, plant materials to which the cat responded positively were offered last during subsequent exposures on other days. Observing a positive response after a negative one demonstrated that environmental (e.g., distractions), mental (e.g., fear) and physical (e.g., disease) conditions did not prevent the cat from responding positively. This too was done to prevent scoring false negative responses. All cats were exposed to the plant materials in their normal living environment, and only after a period of at least ten minutes during which the cats could acclimate to the presence of the investigator. Each cat was given the opportunity to experience the plant materials without interference or intimidation from other cats and in the absence of other possible distractions, which sometimes required retesting cats at other moments. A schematic overview of a typical exposure experiment is presented as Fig 3.
Ninety-eight of the 100 cats were scored live by the same researcher (SB), who knew which plant materials were being offered to the cats at each exposure. Two cats living in a private home were scored (by SB) based on video recordings. Cats showing the characteristic catnip response (sniffing, licking, shaking their head, rubbing (chin/cheek) and rolling on their back, sometimes accompanied by drooling and raking (Table 1)) to a plant material were scored as positive for that plant material. Each positive response was scored as either intense/complete (i.e., sniffing and licking, as well as chin/cheek rubbing or rolling over, for >10 s) or mild/partial (i.e., sniffing and licking only  for >15 s non-stop, or an intense/complete response for <10 s). Cats’ responses were scored as negative when none of at least two exposures resulted in a positive response to the plant material. Negative responses of cats present in the same room but not exposed were not recorded.
Exposure to various parts of silver vine
To study the response of cats to various parts of the silver vine plant in more detail, nine randomly selected domestic cats (three males and six females, aged 8 to 60 months; three domestic medium haired and six domestic short haired) at the MRSPAC were exposed on multiple days to the following six plant materials, all contained within socks: 20 g dried silver vine fruit galls, 20 g dried normal silver vine fruit, 1.5 g powder of dried normal silver vine fruit, sun-dried and fresh leaves (2 and 5 g, respectively) from a three year old silver vine plant not bearing fruit galls (Forestfarm Nursery, Williams, OR, USA) and 15 g woodchips made from commercially available silver vine wood sticks (Duobi Pet Supplies, Hangzhou, China) (Table 2). To generate the powder, two gram of dried normal silver vine fruit was ground using a Mixer Mill 400 (Retsch, Newtown, PA, USA) at 25 cycles per second for three minutes. In all experiments, contact time between the cat’s face and the sock was recorded. The plant material was removed after five minutes of indifference, or after the cat stopped actively engaging with the sock for one minute.
Other Felidae: tigers and bobcats
Nine tigers (Panthera tigris) (seven females and two males, age 16 to 20 years, with age unknown for three of them) and five bobcats (Lynx rufus) (three females and two males, age 8 to 23 years) living at Big Cat Rescue, a sanctuary for exotic cats, in Tampa, FL, USA were included in this study. Twenty gram of catnip (Frontier) or 1.5 g of silver vine powder (Smack) was offered to the tigers inside a folded paper bag (50 × 47 × 30 cm) with small holes made in it (n = 6) or uncontained (n = 3) by a Big Cat Rescue staff member (LB) who was certified to hand out enrichment to cats of all sizes. Tigers were first offered catnip, followed by silver vine five minutes later. If a tiger responded positively to the catnip, silver vine was offered at least 5 min after the tiger had lost interest in the catnip. In the absence of a positive response to catnip, silver vine was offered to the tigers without first removing the catnip, which, in some cases, resulted in exposing them to the silver vine within a couple of meters of the catnip. All bobcats received catnip (n = 1) or silver vine powder (n = 4) in crumpled up paper bags (13 × 8 × 25 cm) with no holes made in them. The plant materials were placed within 0.5 meter of the cat’s face, without disturbing them, so that all animals were aware of the presence of the materials. Minimum possible exposure to each plant material was five minutes for each cat. The cats were not tested between 11:00 and 16:00, because of their low level of activity during these hours.
Dried catnip (leaves and flowers), dried normal silver vine fruit, dried silver vine fruit galls, dried valerian root and Tatarian honeysuckle wood, taken from the batches used for the responsiveness studies, were analyzed using gas chromatography coupled with mass spectrometry to determine concentrations of five known or claimed cat-stimulating compounds isolated from catnip and silver vine: cis-trans nepetalactone [12, 16, 22, 23] (CAS Registry Number 21651-62-7), trans-cis nepetalactone [16, 22, 23] (17257-15-7), actinidine [17, 18] (524-03-8), iridomyrmecin [17–19] (485-43-8) and isodihydronepetalactone  (17672-96-7). For extractions, 0.5 g of each plant material was extracted with 3 ml of dichloromethane while stirred at room temperature for 7 days. The solutions were filtered through a short silica gel column and concentrated by evaporating the solvent at room temperature to 1.5 ml. One hundred μl tridecyl acetate (1.001 g/l) was added to each sample as internal standard to allow for quantification of the compounds and comparison between samples. Extracts were analyzed by gas chromatography coupled with a mass selective detector (GC-MS, GC 7890A/MSD 5975C, Agilent Technologies, Santa Clara, CA, USA). The GC-MS system was equipped with an HP-5 ms fused silica capillary column (30 m, 0.22 mm internal diameter, 0.25 μm film (Agilent Technologies)). Conditions were as follows: inlet pressure: 67 kPa, He-flow: 1.2 ml/min, injector: 250 °C, transfer line 300 °C, electron energy 70 eV. The GC oven temperature was kept at 50 °C for 5 min, followed by raising the temperature with 5 °C/min to 320 °C. Identification of compounds was performed by comparison of their mass spectra and retention indices (determined from a homologous series on n-alkanes (C8-C32)) to those of reference compounds (nepetalactone) and commercial mass spectral libraries (actinidine, iridomyrmecin and isodihydronepetalactone) (Wiley 7, NIST 08). The amount of compound is expressed per gram of dried plant material. The lower limit of detection under these conditions was 0.2 μg per gram.
Statistical analysis was performed using GraphPad Prism version 7.02 (GraphPad Software Inc., La Jolla, CA, USA). A P-value < 0.05 was considered statistically significant. Overlaps between responses to the various plant materials were calculated and plotted as a Venn diagram in R (version 2.7.0) using a custom script from Dr. Thomas Girke (University of California, Riverside, CA, USA).