Microcalorimetric estimation of bacteria in milk

The rapid detection of bacterial contamination is of practical importance to all organizations handling large quantities of milk. Dye reduction tests are frequently used but they require considerable incubation time unless the milk is heavily contaminated, and they have other disadvantages, whereas a sufficiently sensitive calorimeter would speedily indicate the level of bacterial metabolism. The production of heat by bacteria varies widely (see, e.g., Bayne-Jones & Rhees, 1929; Forrest, 1969) but, taking.one of the lower values observed by Bayne-Jones & Rhees we have cal/h/cell of Staphylococcus aureus. This is equal to 5 p cal ml-l min-l in a culture containing 300,000 bacteria/ml. The sensitivity of the flow-through version of a microcalorimeter is of about this value, i.e. 6 p cal/min (Monk & Wadso, 1968). However, many of the bacteria likely to be present in milk produce acid and the heat of neutralization will be added to the heat of metabolism. Thus, it was interesting to use a microcalorimeter (LKB Instruments Ltd., 232 Addington Road, South Croydon, Surrey) to compare the heat production of a sample of milk during the growth of its natural flora, with the number of bacteria determined by the plate count, and with the methylene blue reduction time. A sample of bulked single herd milk was incubated at 25°C during three days and kept in ice during the intervening nights. From time to time a sub-sample was diluted and plated on ‘Yeastrel’ milk agar, a second sub-sample was subjected to the methylene blue test with incubation at 37°C and a third sub-sample was pumped through the calorimeter. The calorimeter contains a heat exchanger by which the temperature of the inflowing liquid is brought to that of a metal heat sink surrounding the measuring cell and the thermocouples. From the heat exchanger the sample passes to the measuring cell and any heat produced in the sample while it is in the cell passes through the thermocouples into the surrounding heat sink. The amplified signal from the thermocouples is recorded and compared with that produced from a known quantity of electricity passed through a known resistance in the measuring cell under the same conditions. In this experiment, sterile separated milk was pumped continuously through the calorimeter between the sampling periods, to provide a base line and during electrical standardization. Previous comparison had shown no heat effect when fresh pasteurized whole milk was pumped through in place of sterile separated milk. It seemed unlikely therefore that a serious error would arise from the use of separated milk to provide a base line against which heat production in full-cream milk could be measured. The results are shown in Fig. 1 from which it is clear that the milk had a basal metabolism of about