The measurement of erythrocyte transketolase activity on a discrete analyser

Transketolase (TK, EC.2.2.1.1) is a thiamine-dependent enzyme which is widely used for assessing thiamine deficiency. The TK activity in erythrocytes is usually decreased during thiamine deficiency and this activity can be enhanced in vitro by the addition of thiamine pyrophosphate (TPP). The relative increase in enzyme activity, expressed as a percentage, is called the TPP effect. The TK kinetic method proposed by Smeets et al. [1] is relatively simple and specific and is based on the measurement of NADH decrease. Minor modifications ofthis method have been proposed [2 and 3] but the method remains tedious if several specimens are assayed. To further simplify the assay, we have modified the method of Smeets et al. for a discrete enzyme analyser, the Gilford System 5 to give a very simple, economical and convenient method. Method Reagents These are essentially similar to those of Smeets et al., but the preparation of reagents modified to improve their storage is outlined. (1) Tris buffer, 0' mol/1, pH 7"8 at room temperature [RT]. (2) Thiamine pyrophosphate, 0.01mol/1 in tris buffer. Dissolve 46.1mg TPP (Sigma) in 10ml buffer. Store 1.0 ml portions at 20C.

NADH decrease. Minor modifications ofthis method have been proposed [2 and 3] but the method remains tedious if several specimens are assayed. To further simplify the assay, we have modified the method of Smeets et al. for a discrete enzyme analyser, the Gilford System 5 (Gilford Instrument Laboratories Inc., Oberlin, Ohio, USA) to give a very simple, economical and convenient method.

Reagents
These are essentially similar to those of Smeets et al., but the preparation of reagents modified to improve their storage is outlined.
Store at -20C in 5 or 10ml portions (unused portions can be refrozen and reused). Just prior to the assay, the haemoglobin concentration is determined by the cyanmethaemoglobin method. This concentration is then adjusted to 30 g/1 with saline and centrifuged to obtain a haemolysate free of stroma.

Gilford System 5
This is a discrete analyser which consists ofa spectrophotometer with a flow-through thermocuvette, computer/printer, automatic dispenser and specimen transport. There are several operating modes including kinetic pre-programmed and general modes, making the analyser versatile and simple to use.
In the kinetic mode the specimen is placed into cups in the transport carousel. The reagent is placed into the automatic dispenser and the computer is programmed to the requirements of the assay. When in operation, the required volume of reagent is dispensed into the sample cup. After the appropriate lag phase the reaction mixture is aspirated into the thermocuvette where equilibrium is reached, then absorbances are read over a predetermined time and the enzyme activity calculated and printed.

Reagent mixture
The proportions given below are sufficient for nine tests, namely four specimens (TK and TPP effect for each) and one 'blank'.

Sample loadin9
Adjacent pairs of sample cups are used for the TK and TK-TPP test for each patient sample, and these contain 0.5 ml buffer, and 0'45 ml buffer and 50 #1 TPP respectively.
A precision pipette is used to introduce 50/A haemolysate into each pair of sample cups, using the 'wash-out' technique to transfer the haemolysate. This is then stirred with the pipette tip to achieve even mixing of the reagent and haemolysate. Failure to do this sometimes causes erroneous results.
A sample cup containing 0"5 ml buffer and 50 #1 haemolysate is used as the blank and is placed in front of the first test sample. This blank serves only to prime the Gilford System and its result is not used.

Operational mode
The initial absorbance of the Gilford spectrophotometer is adjusted to about -0"8 with water.
The 'enzyme-general mode', test number 25, is used to program the Gilford computer/printer. The

Comparison between two methods
Both methods showed good correlation for the TK assay (figure 1) and for the activated TK assay (figure 2). Additional analytical data on the two tests are given in table 1.  Accuracy in measurement Accuracy was further determined by assaying two specimens having known low (L) and high (H) TK activities and three specimens prepared in the following proportions: (2L+ 1H), (H + L), (1L + 2H). Figure 3 shows the TK activities ofthe above samples and the expected correlation line. It would appear that our modified method on the Gilford analyser is quite accurate.

Precision
The within-day precision determined for the modified Gilford method on samples with 'low' and 'high normal' TK activities is shown in table 2. The day-to-day precision has not been done owing to specimen instability.

Stability of samples during assay
The time taken to assay five specimens (10 tests and one blank)is about 100 min. The stability of samples over this period was assessed on several occasions by placing them at the beginning and at the end of the assay: no loss of enzyme activity was detected. However, delay in excess of 2 h causes a fall in enzyme activity of0" U/g Hb. This occurs when large batch samples are loaded on the sample tray at the same time.

TPP ejfect
The upper limit of the TPP effect is normally taken to be 25G.
From the 51 patients randomly selected for this study, five showed elevated (abnormal) TPP effect by the manual method and only three were abnormal by the automated method. Of the two patients who showed equivocal results, one had a normal TK activity by both methods, but the other had a low TK and TPP effect by the automated method. We were unable to repeat the latter observations to confirm the finding.

Discussion
A semi-automated TK method adapted to the centrifugal fast analyser based on the measurement of glucose-6-phosphate has been proposed [4]. The  completed. There is no loss in enzyme activity when specimens are left in the cups for up to 100 min before analysis. Ifthis time is markedly exceeded there is some loss in enzyme activity. Thus for a large batch ofsay 10 specimens (i.e. 20 tests), we recommend that the specimens be loaded in two or more stages so that they are not left for an unduly long time. To minimize or prevent enzyme deterioration, specimens waiting to be assayed should be stored in a refrigerator.
For the automated assay the optimum lag time, established mainly at room temperature which will allow the TK reaction to proceed at a constant rate, similar to the manual method, was found to be 1600 s, of which the last 300 s is at 37C.
The absorbance of the TK test is high and this is due to the NADH and the haemoglobin present. Only instruments which can offset some of this initial high absorbance will be able to measure the decreasing NADH absorbance change due to the TK assay. Both the Unicam SP8000 and the Gilford spectrophotometers can do this adequately.
Extrapolated from our previous data [5], the range for the TK activity was 0"6 to 1.3 U/glib, while the range for the thiamine pyrophosphate (TPP) effect was 0-25?/o. when performed manually is that it is very demanding on the analyst's time. With a suitable spectrophotometer, such as the Unicam SP8000, four tests can be performed in an analytical run, but despite this the change-over of tests and the calculation of enzyme activity is time-consuming. This is not so for our modified TK method on the Gilford analyser. Specimens can be loaded in the sample cups and left unattended until the test is