The response of the malaria mosquito, Anopheles gambiae, to two ...

Physiological Entomology (2001) 26, 142±148

The response of the malaria mosquito, Anopheles gambiae, to two components of human sweat, ammonia and L-lactic acid, in an olfactometer M . A . H . B R A K S 1 , J . M E I J E R I N K 2 and W . T A K K E N Laboratory of Entomology, Wageningen University, Wageningen, the Netherlands

Abstract. In an olfactometer study on the response of the anthropophilic malaria mosquito Anopheles gambiae s.s. (Diptera, Culicidae) to human sweat it was found that freshly collected sweat, mostly of eccrine origin, was attractive, but that incubated sweat was signi®cantly more attractive than fresh sweat. The behavioural response to L-lactic acid and ammonia, the main constituents of sweat, was investigated. L-lactic acid was attractive at one concentration only (11.11 mM) and removal of the L-lactic acid from the sweat by enzymatic decomposition did not affect the attractiveness of sweat. Ammonia caused attraction over a range of 0.1± 13.4 M on glass slides and at 0.84±8.40 mmol/min in an air stream. It is concluded that: human sweat contains kairomones for host-seeking An. gambiae; ammonia is an important kairomone for this mosquito; and that L-lactic acid is not a prerequisite in the attraction of An. gambiae to sweat. Key words. Ammonia, Anopheles gambiae s.s., lactic acid, host-seeking behaviour, human sweat. Introduction Nocturnally active blood-feeding mosquitoes (Diptera, Culicidae) locate and identify their vertebrate hosts mostly by odour. The olfactory cues or kairomones are contained in the expired air and/or skin emanations. Host-seeking Anopheles gambiae Giles sensu stricto (henceforth termed An. gambiae), being highly anthropophilic, orientate mainly to human skin emanations (Mboera & Takken, 1997; Costantini et al., 1998; Takken & Knols, 1999). The volatiles emanating from the skin originate either from the secretions of skin glands or the skin microorganisms or both. Gland secretions can be divided into water-soluble products, mainly from eccrine sweat glands, and fat-soluble products from the sebaceous and apocrine glands. Recently, Braks et al. (1997) reported attraction of An. gambiae to human sweat collected from volunteers performing physical exercise in a warm humid room. Sweat samples that had been incubated for 2 days at 37°C were attractive while freshly collected sweat samples Correspondence: Dr Willem Takken, Laboratory of Entomology, Wageningen University, PO Box 8031, 6700 EH, Wageningen, the Netherlands. Tel.: + 31 317484652; fax: + 31 317484821; e-mail: [email protected] Present addresses: 1Florida Medical Entomology Research Laboratory, 200 9th St. SE, Vero Beach Fl. 32962, U.S.A. and 2Department of Ecology, Lund University, S-223 62 Lund, Sweden. 142

were not (Braks & Takken, 1999). Skin microorganisms during incubation in the sweat sample (Braks et al., 1999) probably produced the volatiles responsible for the attraction. Thermally induced sweat consists principally of eccrine excretion with L-lactic acid, ammonia, urea and electrolytes as the main components (Noble & Somerville, 1974). L-lactic acid has been shown to play an important role in the hostseeking behaviour of the yellow fever mosquito Aedes aegypti (Linnaeus) (Acree et al., 1968; Geier et al., 1996), but behavioural responses of An. gambiae to L-lactic acid have not been reported. Ammonia has occasionally been mentioned in association with host-seeking Aedes mosquitoes (Rudolfs, 1922; MuÈller, 1968; Geier et al., 1999), but not with An. gambiae. In this study, we investigated the general features of sweat samples of different human individuals to identify the constituent volatiles responsible for the attraction of An. gambiae to sweat (Braks & Takken, 1999; Meijerink et al., 2000). For this, sweat collections of 15 volunteers were pooled before testing to minimize volunteer-speci®c effects of the kind reported by Braks & Takken (1999). The attractiveness of the sweat samples over time was determined to assess the volatility level of components responsible for the difference in attraction to fresh and incubated sweat. The role of two main eccrine sweat components, L-lactic acid and ammonia, in the attraction of An. gambiae to sweat was investigated by determining the change in their concentration in the sweat # 2001

Blackwell Science Ltd

Sweat components as kairomone for Anopheles gambiae during the time of incubation, the dose±response relationships and the effect of the selective removal of L-lactic acid from sweat on its attractiveness. Materials and methods Insects The An. gambiae strain used originated from Suakoko, Liberia (by courtesy of Professor M. Coluzzi, Rome) and was maintained under standard insectary conditions (27 6 1°C, 80 6 5% r.h., LD 12 : 12 h). Adults were kept in 30-cm cubic gauze-covered cages and had access to 6% glucose solution. Females were offered blood from a human arm for 10 min twice a week. Eggs were laid on wet ®lter paper, hatched in water trays and the larvae were fed on Tetraminq ®sh food. Pupae were collected daily from the trays and were allowed to emerge in the adult cages. For the bioassay 4±8-day-old female mosquitoes, which had not received a bloodmeal, were used. Behavioural experiments were performed during the last 4 h of the dark period. Collection and treatment of sweat samples Sweat droplets were collected from the foreheads of 15 Caucasian human volunteers (11 males and four females with ages ranging from 25 to 52 years) with sterile glass Pasteur pipettes and put in glass vials (5 mL). Three mL of sweat was collected from each volunteer. Sweat production was stimulated by physical exercise in a warm humid room (30°C, 90% r.h.). Immediately after collection, 1.5 mL sweat was stored immediately at ± 20°C and the other half was incubated under aerobic conditions at 37°C for 2 days, after which it was also stored at ± 20°C. The pH of the fresh sweat samples collected by the different volunteers ranged from 4.4 to 7.4 (5.81 6 0.96 mean 6 SD) and the pH of the incubated sweat ranged from 6.3 to 9.0 (8.8 6 0.76 mean 6 SD). A pooled sample was composed of subsamples collected by the different volunteers (0.6 mL sweat per volunteer). The pH (micro pHelectrode, Inlab 423, Mettler, Toledo AG, Zurich, Switzerland), L-lactic acid concentration (Lactic Diagnostic Kit Sigma, NR 735, St Louis, MO, U.S.A.), urea and ammonia concentration (indophenol method according to Berthelot; Gips & Wibbens-Alberts, 1968) of the sweat were determined. Selective removal of L-lactic acid from the sweat was achieved by adding 50 mL lactate-oxidase (from Pediococcus spp., 34 U/mg, Sigma, 50 U/0.1 mL buffer) and 100 mL catalase (from bovine liver, 3100 U/mg, Sigma, 50 mg/mL buffer) to 1 mL sweat (after Geier, 1995). The buffer (pH 6.4) contained 0.136 g KHPO3 and 0.278 mL 1N NaOH per 100 mL distilled water. After an incubation period of 2 h at 37°C, all the L-lactic acid of the incubated sweat was oxidized, but it was still present in the fresh sweat (both con®rmed with Lactic Diagnostic Kit). To oxidize the remaining L-lactic acid of the fresh sweat, an extra 20 mL lactate-oxidase and 40 mL catalase # 2001

Blackwell Science Ltd, Physiological Entomology, 26, 142±148

143

was added. After another 30-min incubation period all the lactic acid had been oxidized. The control samples (sham) consisted of sweat or distilled water that were treated similarly, only omitting the lactate-oxidase from the buffer. Preparation of chemicals One gram of L-lactic acid (L-lactic free acid, Sigma L6400) was dissolved in 10 mL ethanol (> 98% pure, Merck) and concentrations from 0.001 to 100 mg/mL (1.11 3 10±5 to 1.11 M) were tested in the olfactometer. The aqueous solution of ammonia was purchased from Merck (13.4 M or 227.8 g/L) and diluted further with distilled water. A series of concentrations of ammonia was tested either in the form of an aqueous solution applied on glass slides (1.3 3 10±4 to 13.4 M, which is equivalent to 2.5 3 10±4, 25%, respectively) or in gaseous form. For the latter, 250 mL of 0.1, 1.3 and 13.4 mM ammonia solution, respectively, were put in a 80 L Te¯on gas sampling bag (SKC Inc, Eighty Four, PA, U.S.A.). Subsequently, the bags were ®lled with 80 L humidi®ed warm air (derived from air supply of olfactometer) at least 15 h prior to the experiment to allow the solution to evaporate. This procedure resulted in ammonia concentrations of 1.0, 10.3 and 103.0 p.p.m. in the bags. The control stimuli of the latter consisted of an equivalent amount of air taken from another gas sampling bag ®lled with air and 250 mL distilled water. Behavioural assay A dual-port olfactometer (modi®ed after Braks & Takken, 1999), consisting of a Perspex ¯ight chamber (1.60 3 0.66 3 0.43 m), was used to study the attractiveness of possible stimuli. Charcoal-®ltered, humidi®ed (65 6 5% r.h.), warm air (27 6 0.5°C) was led via a Perspex mosquito trapping device, which was linked to two ports (diameter 4 cm, 28 cm apart), into the ¯ight chamber with a speed of 20 cm/s. Dim light (1 lux) was produced by one incandescent light bulb (75 W) and was ®ltered and scattered through a screen of yellow cloth hanging 1 m above the ¯ight chamber. The stimuli tested were sweat, L-lactic acid and ammonia. In each trial within an experiment, except the one testing ammonia from gas sampling bags, a standard volume of a stimulus was applied on a sandblasted glass slide (5 3 2 cm) that was placed in the left or right trapping device. A control glass slide with an equivalent amount of sweat or solvent was placed in the opposite trapping device. The standard stimulus volume for sweat was 50 mL and for L-lactic acid or ammonia solution it was 100 mL. From the gas sampling bag, ammonia was pumped with 250 mL air/min (air pump model MG-4, AMETEK; release rate: 8.4 3 10±3 to 8.4 mmol/min) through Te¯on tubes (diameter 7 mm) into the trapping device, where it merged with the main air stream of 23.5 L/min. During a 15 min trial, 30 mosquitoes were released together from a container at the downwind end of the ¯ight chamber and were allowed to choose between one of the two trapping devices. Stimuli were alternated between the trapping devices

144

M. A. H. Braks et al.

with each trial. Each successive trial started with fresh mosquitoes, clean trapping devices and freshly applied odour stimuli except for one experiment testing the attractiveness of sweat over time. Three different sets of experiments were performed: 1. Responses to fresh and incubated sweat and attractiveness of sweat over time. The responses of An. gambiae to fresh and incubated sweat were determined by testing them against an equivalent amount of distilled water or each other (Table 2, experiment 1). The attractiveness of sweat over time was assessed in dual choice tests with fresh and incubated sweat while using the same glass slides during six subsequent trials with a 5 min interval (Fig. 1, experiment 2). 2. Responses to L-lactic acid. First, a concentration series of L-lactic acid in ethanol was tested against ethanol (Table 3, experiment 3). Subsequently, the effect of the selective removal of L-lactic acid on the attractiveness of sweat was investigated by testing lactate-oxidase-treated sweat samples (Table 3, experiments 4 and 5) against sham-sweat samples and sham-water. 3. Responses to ammonia. In these experiments, a concentration series of ammonia was tested either as an aqueous solution applied to glass slides (Table 4, experiment 6) or in gaseous form (Table 4, experiment 7). Statistics The total number of mosquitoes caught in the treatmenttrapping device in six replicates was compared with the total number in the control trapping device using Chi-squared tests.

Fig. 1. Responses of An. gambiae to fresh and incubated sweat over time (asterisks indicate signi®cant differences (P < 0.05) between the fresh and incubated sweat at the time of question).

Table 1. Chemical composition of pooled sweat from 15 volunteers. Parameter

Fresh sweat

Incubated sweat

pH

5.3 32.2 19.6 6.3

8.5 22.2 15.3 49.4

L-lactic

acid (mM) Urea (mM) Ammonia (mM)

of mosquitoes caught in the two trapping devices in the other time intervals (Fig. 1).

Results Chemical composition of pooled sweat Pooling of the sweat samples resulted in a single sample of fresh sweat with pH 5.3 and a single sample of incubated sweat with pH 8.5. The L-lactic acid and urea concentration of the sweat decreased during the incubation by 31.1% and 21.4%, respectively. In contrast, the ammonia concentration of sweat increased more than 7.5 times in the same period (Table 1). Responses to fresh and incubated sweat and the attractiveness of sweat over time Compared with the control, signi®cantly more mosquitoes were caught in the trapping devices baited with fresh (P = 0.004) or incubated (P < 0.001) sweat. When tested against each other, the mosquitoes responded more to incubated sweat than to fresh sweat (P < 0.001) (Table 2). When the attractiveness of sweat over time was studied, the preference for incubated sweat was evident only during the ®rst 20 min. In the fourth time interval, 60±75 min after introduction in the olfactometer, a preference for the fresh sweat over the incubated sweat was found (P = 0.0067). There was no signi®cant difference (P > 0.05) between the numbers

Responses to L-lactic acid The trapping device baited with 11.11 mM L-lactic acid caught signi®cantly (P < 0.001) more mosquitoes than the control (with an equivalent amount of ethanol). The number of mosquitoes caught in the traps baited with any of the other Llactic acid concentrations did not differ signi®cantly (P > 0.05) from the control catches (Table 3, experiment 3). Fresh sweat treated with lactate oxidase that selectively decomposed Llactic acid and sham-treated fresh sweat were signi®cantly more attractive than the water controls (P = 0.024 and P < 0.001, respectively) and when these fresh sweat samples were tested against each other (Table 3, experiment 4) they did not differ in attractiveness (P > 0.05). Incubated sweat treated with lactate oxidase and sham-treated incubated sweat was also signi®cantly more attractive than the control (P < 0.001) and no difference (P > 0.05) was found between the two differently treated incubated sweat samples when tested against each other (Table 3, experiment 5). Responses to ammonia Ammonia (100 mL of 0.1, 1.3 and 13.4 M) applied to glass slides attracted signi®cantly more mosquitoes to the trapping # 2001


Sweat components as kairomone for Anopheles gambiae

145

Table 2. Responses of An. gambiae to sweat (experiment 1). Stimuli

Response

a

N

Experiment

Treatment

Control

Treatment

Control

Chi square

1

Fresh sweat Incubated sweat Incubated sweat

Water Water Fresh sweat

32 93 81

13 6 15

** *** ***

b

c

+

±

n

6 6 6

0 0 0

0 0 0

a

The response is given as the total number of mosquitoes caught in either the treatment or control trapping device. Signi®cant differences (*: P < 0.05, **: P < 0.01 or ***: P < 0.001) or no signi®cant differences (ns: P > 0.05) found between the total number of mosquitoes caught in the treatment and control trapping device (Chi-squared test). c N = number of replicates. The distribution of the preferences for either the treatment (+), the control (±) or neutral (n) in the individual replicates are also shown. b

Table 3. Responses of An. gambiae to L-actic acid (experiment 3) and sweat samples treated with lactate-oxidase (experiments 4 and 5). Stimuli

Response

a

N

Experiment

Treatment

Control

Treatment

Control

Chi square

3

0.01 mM LA d 0.11 mM LA 1.11 mM LA 11.11 mM LA 0.11 M LA 1.11 M LA Fresh sweat + LOe Fresh sweat sham Fresh sweat + LO Incubated sweat + LO Incubated sweat sham Incubated sweat + LO

Ethanol Ethanol Ethanol Ethanol Ethanol Ethanol Water sham Water sham Fresh sweat sham Water sham Water sham Incubated sweat sham

33 31 32 48 28 39 41 53 41 69 60 65

39 31 28 14 40 25 23 18 47 16 14 49

ns ns ns *** ns ns * *** ns *** *** ns

4 5

b

c

+

±

n

3 4 2 5 3 3 5 6 1 6 5 5

3 2 2 0 3 3 0 0 4 0 1 1

0 0 2 1 0 0 1 0 1 0 0 0

a/b/c d

see description under Table 2. LA = L-lactic acide LO = lactate oxidase.

Table 4. Responses of An. gambiae to ammonia from glass slides (experiment 6) or gas-sampling bags (experiment 7). Stimuli

Response

a

N

Experiment

Treatment

Control

Treatment

Control

Chi square

6

0.1 mM NH3 1.3 mM NH3 13.4 mM NH3 0.1 M NH3 1.3 M NH3 13.4 M NH3 0.08 mmol NH3/min 0.84 mmol NH3/min 8.40 mmol NH3/min

Water Water Water Water Water Water Clean air Clean air Clean air

31 35 35 69 63 48 26 70 60

22 24 33 15 21 17 21 26 22

ns ns ns *** *** *** ns *** ***

7

b

c

+

±

n

4 3 2 6 6 5 5 5 4

1 2 1 0 0 0 0 1 2

1 1 3 0 0 1 1 0 0

a/b/c d

see description under Table 2. NH3 = ammonia.

device than the control (P < 0.001) (Table 4, experiment 6). Also, ammonia released at a rate of 0.84 or 8.40 mmol/min from the gas sampling bags attracted signi®cantly (P < 0.001) more mosquitoes to the traps than the control air¯ow. The # 2001


catches with 100 mL of 0.1 ±13.4 M ammonia solution or ammonia at a rate of 0.08 mmol/min from the gas sampling bags (Table 4, experiment 7) did not differ from the control (P > 0.05).

146


Discussion Responses to fresh and incubated sweat and the attractiveness of sweat over time Anopheles gambiae was attracted to sweat pooled from 15 individuals and on all occasions incubated sweat was preferred to fresh sweat when tested simultaneously. In a separate study we have observed behavioural and electrophysiological responses of An. gambiae to pooled incubated sweat (Meijerink et al., 2000) and it appears that incubated sweat is an attractive stimulus for this mosquito in the laboratory. This positive effect of incubation is probably due to the action of skin microorganisms, which can be collected from the skin in aqueous skin secretions (Braks & Takken, 1999). However, in contrast to our previous ®ndings with individual sweat samples (Braks & Takken, 1999; Meijerink et al., 2000), in the present study pooled fresh sweat was preferred to the control (water). As pooling of sweat minimizes volunteer-speci®c effects we suggest that some attraction to fresh sweat should be considered the general rule, even though our previous data from three individuals included some exceptions to this rule. This result suggests that kairomones to which An. gambiae responds were also present in fresh sweat but that the quantity or quality of the attractive volatiles was enhanced strongly during incubation. Skin microorganisms are presumed to break down sweat-borne compounds into smaller, more volatile components (Braks et al., 1999). This, together with the fact that the preference of the mosquitoes for the incubated sweat was lost within 20 min after exposure in the olfactometer, suggests that the components responsible for the preference of An. gambiae for incubated sweat to fresh sweat are highly volatile. Response to L-lactic acid The sweat samples contained L-lactic acid within the range 4.0±40 mM, as reported by Noble & Somerville (1974). The observed reduction of L-lactic acid during incubation was probably due to the utilization of this component by skin microorganisms (Bergeim & Cornbleet, 1943; Smith, 1971). Only one out of the six concentrations of L-lactic acid tested was attractive for An. gambiae, namely 11.11 mM. The amount at the source (1.1 mmol/100 mL) was close to the amount of Llactic acid present in the incubated sweat (1.1 mmol/50 mL). Attraction to a broader range of L-lactic acid concentrations (2.2±222.2 mM) has been shown for Ae. aegypti (Geier et al., 1996; Geier & Boeckh, 1999). The latter authors showed that the attractiveness of skin washings for Ae. aegypti was completely lost after the selective removal of L-lactic acid. However, skin washings containing L-lactic acid induced a higher response in this mosquito than L-lactic acid alone, indicating a strong synergistic effect of L-lactic acid with other components present in the skin washings (Geier et al., 1999). In contrast, in our study the selective removal of lactic acid from the sweat did not affect its attractiveness for An. gambiae, indicating that L-lactic acid is not a prerequisite in the

attraction of this species to sweat. In a recent study, Healy & Copland (2000) also did not ®nd a behavioural response to a natural concentration of lactic acid, giving further support to our ®ndings. The insigni®cance of lactic acid in the response of An. gambiae to the sweat samples had previously been suggested following arti®cial alkalization of sweat to pH 10 that reduces volatilization of L-lactic acid (Braks et al., 1997). The differences between these studies with Ae. aegypti and An. gambiae indicate that either L-lactic acid is a stronger stimulus for Ae. aegypti than for An. gambiae or that the responses to sweat and skin washings are not based on the same set of volatiles. The former suggests differences in odour-mediated host-seeking behaviour between the two species that are conceivable. The latter is most likely considering the different procedures followed for the collection of the samples of sweat or skin washings and the differences in attractiveness of the stimuli over time. The skin washings were made by rubbing hands and feet with cotton wool soaked in methanol, which were left to dry for a day at room temperature. The skin washings proved to be attractive for Ae. aegypti for more than 2 months after application to ®lter paper, and storage at room temperature (Geier, 1995). This suggests that most of the attraction to skin washings was not based on highly volatile components. Our procedure of placing the sweat in airtight vials immediately after collection prevented the loss of (highly) volatile components either present in the sweat or produced during incubation. Response to ammonia To our knowledge, this is the ®rst report showing that An. gambiae responds to ammonia. Our investigation of ammonia was initiated mainly after recognizing that the pH of the sweat samples increased during incubation (Braks & Takken, 1999). It seemed likely that this was due to the production of ammonia from urea and amino acids by microorganisms present in the samples (Bergeim & Cornbleet, 1943; MuÈller, 1968). A rise in pH was observed in the sweat of all volunteers during incubation in sealed containers, accompanied by an increase of the ammonia concentration and a decrease of the urea concentration. As more ammonia was formed than could be explained by the metabolism of urea, it is likely that other nitrogen-components, such as amino acids, present in sweat (Gitlitz et al., 1974) must have been deaminated. From the results of the experiments with varying concentrations of ammonia, we deduce that the behavioural threshold of ammonia lies between 100 mL of 13.4 mM and 0.1 M. The amount of ammonia in the incubated sweat at the source, 50 mL of 49.4 mM, lies within this transition interval; but the ammonia in the fresh sweat, 50 mL of 6.3 mM, lies below this interval. The results from tests in which ammonia was released from gas-sampling bags demonstrate a clear response of An. gambiae to a low but constant ¯ow of ammonia ranging from 0.84 to 8.40 mmol/min. These odorous air ¯ows were diluted even further (~ 100 times) by the main stream of clean air. From these responses to ammonia and the observed differences in ammonia concentration between fresh and incubated sweat, # 2001


Sweat components as kairomone for Anopheles gambiae as well as the decline in attractiveness of the sweat within 20 min of unsealing it, we suggest that ammonia contributes to the differential attractiveness of fresh and incubated sweat. The responses to fresh sweat, which contains only small amounts of ammonia, indicate that additional components must also play a role in the attraction of An. gambiae to sweat. These may include short-chain aliphatic carboxylic acids, as a mixture of these was found to be attractive for An. gambiae (Knols et al., 1997) and individual acids evoked electrophysiological responses (Meijerink & van Loon, 1999). In addition, electrophysiological responses to other volatile components detected in sweat are reported by Meijerink et al. (2000). Rudolfs (1922) recognized ammonia, together with carbon dioxide, as the ultimate products of decomposition of the human body and reported positive behavioural responses of Ae. sollicitans and Ae. cantator to these compounds. MuÈller (1968) found that ammonia was not attractive for Ae. aegypti but that it increased the activity of this mosquito. A similar response was found with trunk sweat that had been incubated for 24 h. He suggested that the latter attraction was caused by the large amounts of ammonia present in the incubated sweat. Recently, Geier et al. (1999) showed that ammonia increases the attraction of Ae. aegypti to lactic acid. A mixture of ammonia, lactic acid and two fatty acids, was the most attractive arti®cial odour blend for Ae. aegypti to date (Bosch et al., 2000). However, when offered alone, ammonia was not attractive for this mosquito species. Responses to ammonia associated with host-location have also been reported from other haematophagous arthropods, e.g. ticks (El-Ziady, 1958; Haggart & Davis, 1980; Sonenshine et al., 1986) and horse ¯ies (Hribar et al., 1992). In addition, sensitivity to ammonia has been reported for insects such as haematophagous bugs (Taneja & Guerin, 1995, 1997) and body lice (Mumcuoglu et al., 1986), where ammonia functions as an aggregation pheromone and is derived from the faeces of conspeci®cs. In fruit ¯ies it is an oviposition cue (Mazor et al., 1987). Recent research showed that relatively large amounts of ammonia (3.36 6 2.08 mmol/30 min) are lost with sweat during intense exercise and this probably has the important function of preventing excessive increase of ammonia levels in the blood (Czarnowski & GoÂrski, 1991). At rest, the loss of ammonia through sweating is probably small, as the concentration of ammonia in blood is extremely low (~35 mmol/L, Czarnowski & GoÂrski, 1991) and the rate of sweating is low (~0.5 L sweat per 24 h under basic metabolic conditions, Walter, 1972). Nonetheless, this brings about an accumulation of non-volatile sweat components, such as urea, on the skin (Jenkinson, 1980). Urea, together with other nitrogen-rich compounds, is utilized by skin microorganisms, resulting in the emission of ammonia as an ultimate breakdown product (Bergeim & Cornbleet, 1943). The emanation rate from the skin of ammonia, either of physiological or bacterial origin, or both, is not known. Although ammonia is not restricted to human emanations, the level of ammonia emission may be human-speci®c. Haggart & Davis (1980) suggested that the two types of ammonia-sensitive neurones found in ticks encode for different levels of ammonia in the environment, one responding to the # 2001


147

low ammonia levels as found in sweat and the other responding to high levels of ammonia produced from urine and faeces along an animal trail. Recently, electrophysiological studies showed that neurones on the antennae of An. gambiae that responded to incubated sweat are also sensitive to ammonia (Meijerink et al., 2001). We conclude that ammonia is an important kairomone for host-seeking An. gambiae and that L-lactic acid is not a prerequisite in the attraction of An. gambiae to sweat. Acknowledgements The 15 volunteers are acknowledged for kindly providing the sweat samples. M. van Helden is thanked for his initial suggestions. We are also grateful to our colleagues at the Laboratory of Entomology and especially W. Ament and V. Schreurs for helpful discussions. J.J.A. van Loon is thanked for helpful comments on an earlier draft of the manuscript. Our thanks are also due to R. Koopmanschap for his part in the ammonia analyses. We thank F. van Aggelen, A. Gidding and L. Koopman for rearing the mosquitoes. The research was ®nanced by The Netherlands Organization for Scienti®c Research under reg. no. 805-33.043/P. References Acree, F. Jr, Turner, R.B., Gouck, H.K., Beroza, M. & Smith, N. (1968) L-Lactic acid: a mosquito attractant isolated from humans. Science, 161, 1346±1347. Bergeim, O. & Cornbleet, T. (1943) The antibacterial action of the lactic acid and volatile fatty acids of sweat. American Journal of Medical Science, 205, 785±792. Bosch, O.J., Geier, M. & Boeckh, J. (2000) Contribution of fatty acids to olfactory host ®nding of female Aedes aegypti. Chemical Senses, 25, 323±330. Braks, M.A.H., Cork, A. & Takken, W. (1997) Olfactometer studies on the attraction of Anopheles gambiae sensu stricto (Diptera: Culicidae) to human sweat. Proceedings of Experimental and Applied Entomology, NEV, Amsterdam, 8, 99±104. Braks, M.A.H., Knols, B.G.J. & Anderson, R.A.A. (1999) The role of human skin micro¯ora and Plasmodium malaria parasites in mosquito±host interactions. Parasitology Today, 15, 409±413. Braks, M.A.H. & Takken, W. (1999) Incubated human sweat but not fresh sweat attracts the malaria mosquito. Anopheles gambiae sensu stricto. Journal of Chemical Ecology, 25, 663±672. Costantini, A., Sagnon, N., Della Torre, A., Diallo, M., Brady, J., Gibson, G. & Coluzzi, M. (1998) Odor-mediated host preference of West African mosquitoes, with particular reference to malaria vectors. American Journal of Tropical Medicine and Hygiene, 58, 56±63. Czarnowski, D. & GoÂrski, J. (1991) Sweat ammonia excretion during submaximal cycling exercise. Journal of Applied Physiology, 70, 371±374. El-Ziady, S. (1958) The behaviour of Ornithodoros erraticus (Lucas, 1849), small form (Ixodoidea, Argasidea) towards certain environmental factors. Annals of the Entomological Society of America, 51, 317±336. Geier, M. (1995) Verhaltensversuche mit Gelb®ebermuÈcken Aedes aegypti zur AufklaÈrung des attractiven Reizmusters bei der

148


olfaktorischen Wirts®ndung. Dissertation UniversitaÈt Regensburg, Germany. Geier, M. & Boeckh, J. (1999) A new Y-tube olfactometer for mosquitoes to measure the attractiveness of host odours. Entomologia Experimentalis et Applicata, 92, 9±19. Geier, M., Bosch, O.J. & Boeckh, J. (1999) Ammonia as an attractive component in host odour for the yellow fever mosquito, Aedes aegypti. Chemical Senses, 24, 647±653. Geier, M., Sass, H. & Boeckh, J. (1996) A search for components in human body odour that attract females of Aedes aegypti. Olfaction in Mosquito±Host Interactions (ed. by G. R. Bock and G. Cardew), pp. 132±148. Ciba Foundation Symposium, no. 200. Gips, C.H. & Wibbens-Alberts, M. (1968) Ammonia determination in blood using the TCA direct method. Clinica Chimica Acta, 22, 183± 186. Gitlitz, P.H., Sunderman, F.W. Jr & Hohnadel, D.C. (1974) Ionexchange chromatography of amino acids in sweat collected from healthy subjects during sauna bathing. Clinical Chemistry, 20, 1305±1312. Haggart, D.A. & Davis, E.E. (1980) Ammonia-sensitive neurones on the ®rst tarsi of the tick, Rhipicephalus sanguineus. Journal of Insect Physiology, 26, 517±523. Healy, T.P. & Copland, M.J.W. (2000) Human sweat and 2oxopentanoic acid elicit a landing response from Anopheles gambiae. Medical and Veterinary Entomology, 14, 195±200. Hribar, L.J., Leprince, D.J. & Foil, L.D. (1992) Ammonia as an attractant of adult Hybomitra lasiophthalma (Diptera: Tabanidae). Journal of Medical Entomology, 29, 346±348. Jenkinson, D.M. (1980) The topography, climate and chemical structure of the mammalian skin surface. Proceedings of the Royal Society of Edinburgh, 79B, 3±22. Knols, B.G.J., Van Loon, J.J.A., Cork, A., Robinson, R.D., Adam, W., Meijerink, J., De Jong, R. & Takken, W. (1997) Behavioural and electrophysiological responses of the female malaria mosquito Anopheles gambiae s.s. (Diptera: Culicidae) to Limburger cheese volatiles. Bulletin of Entomological Research, 87, 151±159. Mazor, M., Gothilf, S. & Galun, R. (1987) The role of ammonia in the attraction of females of the Mediterranean fruit ¯y to protein hydrolysate baits. Entomologia Experimentalis et Applicata, 43, 25± 29. Mboera, L.E.G. & Takken, W. (1997) Carbon dioxide chemotropism in mosquitoes (Diptera: Culicidae) and its potential in vector surveillance and management programmes. Review of Medical and Veterinary Entomology, 85, 355±368.

Meijerink, J., Braks, M.A.H., Brack, A., Adam, W., Dekker, T., Posthumus, M.A., Van Beek, T.A. & Van Loon, J.J.A. (2000) Identi®cation of olfactory stimulants for Anopheles gambiae from human sweat samples. Journal of Chemical Ecology, 26, 1367± 1382. Meijerink, J., Braks, M.A.H. & Van Loon, J.J.A. (2001) Olfactory receptors on the antennae of the malaria mosquito Anopheles gambiae are sensitive to ammonia and other sweat-borne components. Journal of Insect Physiology, 47, 455±464. Meijerink, J. & Van Loon, J.J.A. (1999) Sensitivities of antennal olfactory neurons of the malaria mosquito, Anopheles gambiae, to carboxylic acids. Journal of Insect Physiology, 45, 365±373. MuÈller, W. (1968) Die Distanz- und Kontakt-Orientierung der StechmuÈcken (Aedes aegypti) (Wirts®ndung, Stechverhalten und Blutmalzeit). Zeitschrift fuÈr Vergleichende Physiologie, 58, 241± 303. Mumcuoglu, Y., Galun, R. & Ikan, R. (1986) The aggregation response of the human body louse (Pediculus humanus) (Insecta: Anoplura) to its excretory products. Insect Science and its Application, 7, 629± 632. Noble, W.C. & Somerville, D.A. (1974) Microbiology of Human Skin. W.B. Saunders Company Ltd, London. Rudolfs, W. (1922) Chemotropism of mosquitoes. Bulletin of the New Jersey Agricultural Experimental Station, 367, 4±23. Smith, R.F. (1971) Lactic acid utilization by the cutaneous Micrococcaceae. Applied Microbiology, 21, 777. Sonenshine, D.E., Taylor, D. & Carson, K.A. (1986) Chemically mediated behaviour in acari: adapations in ®nding hosts and mates. Journal of Chemical Ecology, 12, 1091±1108. Takken, W. & Knols, B.G.J. (1999) Odor-mediated behavior of Afrotropical malaria mosquitoes. Annual Review of Entomology, 44, 131±157. Taneja, J. & Guerin, P.M. (1995) Orientated responses of the triatomine bugs Rhodnius prolixus and Triatoma infestans to vertebrate odours on a servosphere. Journal of Comparative Physiology A, 176, 455±464. Taneja, J. & Guerin, P.M. (1997) Ammonia attracts the haematophagous bug Triatoma infestans: behavioral and neurophysiological data on nymphs. Journal of Comparative Physiology A, 181, 21±34. Walter, W.G. (1972) Warmteregulatie. Grondslagen der Neurofysiologie, pp. 27. Elsevier, Amsterdam. Accepted 31 January 2001

# 2001
