TRANSACTIONS OF THE ROYAL SOCIETY OF TROPICAL MEDICINE AND HYGIENE. Vol. 58. No. 4. July, 1964.

363

NOTES ON THE FATE OF AVIRULENT P A S T E U R E L L A P E S T I S IN FLEAS

BY

LEO KARTMAN

(from the San Francisco Field Station, C.D.C., U.S. Public Health Service, California)

AND

The Late S T U A R T F. Q U A N

We have demonstrated previously by the use of in vitro feeding techniques (KARTMAN,

1954) that an avirulent strain of Pasteurella pestis (A1122) is capable of producing ventricular masses and proventricular blockage in fleas, Xenopsylla spp, much in the manner of virulent strains of the plague bacillus (QuAN, KARTMAN and MCMANuS, 1954; and KARTMAN, QUAN, and MCI~IANUS, 1956). In this connection we have additional observations that should become part of the published record.

Details of the artificial feeding methods and the bacteriological procedures are given in the above cited papers. Fleas of both sexes (X. cheopis) were fed upon six avirulent strains of P. pestis in suspensions of Rattus norvegicus blood held at approximately 37°C. After feeding, each flea was examined to determine whether it was blooded and it then was placed upon an albino rat, a separate rat being used as host for each batch of fleas infected with each strain of P. pestis. At different intervals, fleas were removed from the rats, ex- amined for visual evidence of P. pestis multiplication, and then each flea was triturated and cultured for bacteriological evidence of infection. Plate counts of the number of P. pestis per flea also were made at appropriate intervals. Only those fleas recovered from the rats were included in the data.

The data in Table I indicate that strain A1122 tended toward the same rate of blocking as determined in our previous experiments (see citations above). The strains "14" and B741-10-9 gave similar results, showing a capacity to develop and block the flea host. Al- though dark masses were noted in the stomachs of fleas fed some of the other strains, none of these apparently was able to maintain itself in the fleas and produce convincing infection, not to mention proventricular blockage. The average plate counts of P. pestis in flea samples at 12 and 120 hours after the infectious blood meal were respectively as follows for the several strains used: Al122, 213x10 z an:l 190xl0Z; "14," 850x10 z and 405x10Z; Pp52-11, 2x103 and 0; EV76-20, 23x10 z and 0; "TRU" , 75x10 z and 0; B741-10-9, 166x10 z and 97x10 z. In the case of strains Al122, "14," and B741-10-9 the plate counts over a period of 12 days confirmed our previous observations that, following a decline during the first few days, the bacilli multiplied gradually but surely in the flea host.

We had speculated that "smooth" strains like Al122 could produce large masses of P. pestis in the flea's ventriculus and proventriculus, and eventually block the proventricular valve since the degree of smoothness is associated with the proteinaceous envelope of the bacillus and thus may determine how "sticky" a clump of bacteria will be. On the other hand, a completely rough strain like " T R U " is excreted easily and the flea rapidly clears

364

TABLE I.

N O T E S O N T H E F A T E O F A V I R U L E N T Pasteurella pestis I N F L E A S

The development in X. cheopis of avirulent strains of P. pestis during a period of 26 days after the infectious blood meal fed in vitro.

Pasteurella pestis

Strain

Al122

Pp52-11

EV76-20

"14"

" T R U "

No./c.c.

2 . 4 x 109

1.2 x 109

0.8 x 109

2 .0x109

2.0 x 10 9

B741-10-9 1.7 x 109

Number of fleas

Used*

148

122

59

55

103

49

In~cted

No. %

119 80.4

14 11.4

16 27.1

33 60.0

6 5.8

11 22.4

Fleas with massest

No. i %

113 76.3

8 6.5

11 18.6

31 56.0

1 0.9

9 18.3

Fleas blocked~

No. %

43 36.1

0 0

0 0

! 9 27.2

! o o

l 2 1 8 . 1

* Fleas included are only those recovered from albino rats and cultured (see text). t Per cent. showing ventricular masses of total fleas used. :~ Per cent. of infected fleas.

itself. Only further work can elucidate the nature of this phenomenon which may have some bearing on the ability of virulent strains of P. pestis to develop in the flea host.

In the course of these experiments, and others, we noted a tendency toward different blocking rates in fleas fed varying concentrations of strain Al122. To test this hypothesis, seven different concentrations of strain Al122 in rat blood were fed to fleas. The fleas then were maintained individually in test tubes for from 15 to 30 days; they were fed on a mouse periodically and examined for proventricular blockage. Results are shown in Table II. These data suggest a direct correlation between the number of bacilli in the infectious meal and the blocking rates in the fleas. Thus the degree of bacteraemia in natur- ally infected wild rodents may be one of the limiting factors that determine the number of fleas becoming infective. Conversely, in tests of the vector efficiency of fleas, it would be of interest to test the hypothesis that different species of fleas may have different "blocking thresholds" when the number of P. pestis ingested is constant.

TABLE II. The rate of proventricular blocking in X. cheopis after in vitro feeding on various concen- trations of avirulent P. pestis strain Al122.

Finding

No. fleas blocked/No, fed

Per cent. blocked

1 : 1

6/28

21.4

Dilution of P. pestis in infectious blood meals*

1 : 5 1 : 2 5 1 : 125 [ 1 : 6 2 5 1 :3125 I

9/24 7/19 5/24 z, 2/18 1/19 P

37.5 36.8 20.8 t 11.1 5.3

1 : 15625

o/18

0

* The approximate number of bacilli in millions per c.c. were respectively: 1000-2000, 200-400, 40-80, 8-14, 1-3, 0.3, 0.06.

LEO KARTMAN AND STUART F. QUAN 365

SUMMARY

By the use of in vitro feeding techniques, fleas (XenopsyUa cheopis) were fed upon six avirulent strains of PasteureUa pestis in suspensions of rat blood. Three of the strains (Al122, "14," and B741-10-9) showed a capacity to multiply in the fleas and eventually to cause blockage of the proventricular valve. The other strains apparently did not survive in the fleas. Plate counts of the number of bacilli of each strain in fleas, processed at inter- vals after the infectious blood meal, confirmed the findings with regard to blocking rates.

In another experiment, seven different concentrations of avirulent P. pestis strain A1122 were fed to X. cheopis. The results showed a direct relation between the number of bacilli in the infectious blood meal and the blocking rates in the fleas. Thus the vector efficiency of fleas may be limited by a "threshold" of the degree of bacteraemia in wild rodents infected with P. pestis

REFERENCES

KARTMAN, L. (1954). Exp. Parasit., 3, 525. , QUAN, S. F. & McMArCtlS, A. G. (1956). 1bid., 5, 435.

QUAN, S. F., KAaTMAN, L. & McMANuS, A. G. (1954). Science, 120, 1101.