Clin. Biochem. 9, (4) 192-194 (1976)

A Comparative Study of Micro.Ammonia Determinations in Plasma Using Two Different Methods

ERIC J. SAMPSON and LAURENCE M. DEMERS

Division of Clinical Pathology, The Milton S. Hershey Medical Center, The Pennsylvania State University, Hershey, Pennsylvania 17033, USA

(Accepted March 30, 1976)

CLBIA 9, (4) 192-194 (1976) Clim Biochem.

Sampson, Eric J. and Demers, Laurence M.

Division of Cli~zical Pathology, The Milto?z S. Hershey Medical Center, The Pennsylvania State University, Hershey, Pa 17033 USA.

A COMPETITIVE STUDY OF MICRO-AMMONIA DETERMINATIONS IN PLASMA USING TWO D I F F E R E N T METHODS

Two commercially available kits for determining blood ammonia were compared, an enzymatic method requiring a sample volume of 100~l and a nonenzyma- tic method requiring 1 ml of sample. Calibration curves for both kits were linear to 270 ~mol/I, how- ever correlation studies on plasma samples revealed a lack of agreement between methods (r ~ 0.825, Y ---- 0.76X -t- 51). hnproved correlation occurs when portein interference is eliminated from the enzy-matic procedure. Finally, the nonenzymatic method is modi- fied to accommodate 100 ~l of sample and compared to the original method.

I'VIOST METHODS FOR DETERMINING BLOOD AMMONIA have requi red sample volumes in excess of one ml to achieve adequate precision. In newborn infants , this p resen t s a problem i f capi l lary sampl ing is used.

Recently, a commercial ly available ki t for de te rmin- ing blood ammonia levels on 100 /zl of se rum was in- t roduced which employs the enzyme g lu tamic dehydro- genase to catalyze the fol lowing reac t ion :

H ÷ + o~-ketoglutarate -F N A D H + NH.~ ~ L-g lu ta - mate -~- N A D ÷ -+- H..,O. The equi l ib r ium of this reac- tion lies fa r to the r ight , so tha t changes in absorbance at 340 nm are direct ly propor t ional to the ammonia concentra t ion.

Ano the r ki t involves the use of an ion exchange resin which selectively adsorbs ammonia f r o m plasma. The ammonia is then eluted and reacted wi th hypo- chlor i te and phenol to yield an indophenol (Ber the lo t ' s r eac t ion) . This method is comparable to the recom- mended procedure described by Kingsley and Tage r (').

We examined both kits, wi th adaptat ions , for pos- sible use in the de te rmina t ion of ammonia on capil lary samples.

Correspondence: Dr. L. Demers, Department of Pathology, The M. S. Hershey Medical Center, Her- shey, Pennsylvania 17033, USA.

MATERIALS AND METHODS

"Blood Ammonia Test" kit (Hyland, Costa, Mesa, Calif. 92626) included the following reagents: 4 mol/l sodium chloride, phenol color reagent composed of 5 g phenol and 0.55/~ (w/v) sodium nitroferricyanide, a solution of (15 g/ l ) sodium hypochlorite and 178 g / l sodium hydroxide, ion exchange resin, and nitrogen standard 707 mg/l .

"Eskalab Reagent for Determination of Ammonia in Blood" (Smith Kline Instruments, Inc., Palo Alto, Califor- nia 94306) included two tablets labeled "Ammonia" and "Ammonia No. 2". The tablet labeled "Ammonia" contained : 7 ~mol a-ketoglutarate, 0.28 ~mol reduced nicotinamide adenine dinucleotide, 4.2 .~mol adenosine diphosphate, 126 .~mol tris(hydroxymethyl) aminomethane, 28~mol succinic acid, and 48 ~mol potassium bicarbonate. The tablet labeled "Ammonia No. 2" contained 0.42 IU of glutamic dehydrogenase.

Additional reagents included: reagent grade tris base (Schwartz/Mann), perchloric acid (Baker), and distilled deionized water. Disposable plastic containers and MLA pipettes with disposable tips were used to avoid ammonia contamination. Absorbance measurements were performed on a Gilford Model 240 spectrophotometer equipped with a temperature controlled cell block and on a Gilford Model 300 N spectrophotometer, with a flow through cuvette.

Standard ammonia solutions were prepared from the Hyland nitrogen standard. Patient samples were collected in heparinized tubes and the separated plasma allowed to stand at room temperature for several hours prior to running comparative analyses to avoid large changes in ammonia concentration reported to occur shortly after col- lection':' :"

Hyland Macro-Method. The Hyland method was per- formed according to the manufacturer 's instructions. One ml of standard or sample, 2 ml of water, and five drops of resin were added to a 15 ml plastic tube. The tubes were capped and mixed by inversion for five minutes, af ter which the supernatant was removed from the resin and the resin-washed twice with 10 ml portions of water. After thoroughly decanting the final wash, 1 ml of 4 mol/l NaCl and 1 ml of phenol color reagent were added to each tube and allowed to stand for 3 minutes. One ml of alkali hypo- chlorite reagent was then added to initiate the Ber- thelot reaction and the mixture incubated at 37" for 15 minutes. Finally, 3 ml of water were added and the result- ing absorbance measured at 635 nm.

Hyland Micro-Method. The sample and all reagent vol- umes were reduced to 1/10th of the original volume em- ployed with the Hyland Macro Procedure described above. Only 1 drop of resin was used and for the Berthelot reac- tion the mixture was incubated at 56" for 6 minutes. Similarly, the size of the container was reduced to a Tech- nicon T M sample cup (4 ml) to accommodate the smaller volume.

Eskalab Method. This method was performed according to the manufacturer 's instructions: 0.1 ml of standard or sample, 1.4 ml of water, and one "Ammonia" tablet were added to a cuvette. After 10 minutes at room temperature, the absorbance was measured at 340 nm and "Ammonia

BLOOD AMMONIA DETERMINATIONS 193

. 6 t

.5'

.4 ̧

AMMONIA CONCENTRATION (umol/I}

Fig. 1 ~ Compar ison o f cal ibrat ion curves for Eska lab me thod ( • ) Y "- O.-llX ~ 0.005; Mod i f i ed Eska lab me thod w i t h depro te in i za t ion (C]) Y -- 1 .13X -k 0.010; Modi f i ed Eska lab me thod ( A ) Y -~ 1.92X -t- 0.009; H y - la~d macro-me~hod (C)) Y = 2.11X {J.O05.

2 0 0

E =

,1-

ioc ',s

oo <f

uJ

~oc , /~ o, ¥.o.z6X*~l ,~

/ . ~.o.s2s

• 0

" ,~o ~ 2~o o(~ H Y / A N O M A C R O - M E T H O D |umol / I )

Fig. 2 ~ Compar i son o f E s ka l a b me th - od and H y l a n d Macro-method.

No. 2" tablet was dissolved in the cuvette solution. Af ter 10 minutes at room temperature, a second absorbance read- ing was recorded. Modified Eskalab Method. The "Ammonia" tablet was dissolved in 1.5 m[ of water and 0.2 ml of this solution was added to "Ammonia No. 2" tablet to produce a slurry. Then, 0.i ml of s tandard or sample and 0.2 ml of solution containing the dissolved "Ammonia" tablet were added to a micro°cuvette (0.5 ml, 1 cm light path) , incubated at 37 ° for 15 minutes, and the absorbance measured at 340 nm. Finally, 0:02 ml of the s lurry of "Ammonia No. 2" tablet was added, mixed, and a second absorbance read- ing recorded af ter 30 minutes at 37 ° .

Modified Eskalab Method with Deproteinization. The "Ammonia" tablet was dissolved in 0.7 ml of 1 mol/1 t r i s base and a s lurry of "Ammonia No. 2" tablet was prepared with the addition of 0.I ml of water. Then 0.I ml of stand- ard or sample was added to 0.2 ml o~ 0.5 tool/1 CHIO~ and centrifuged at 2000 rpm for 10 minutes. Af te r centr ifuga- tion, 0.2 ml of the supernatant was combined with 0.1 ml of solution containing the dissolved "Ammonia" tablet in a microcuvette, incubated at 37 ° for 10 minutes, and the absorbance measured at 340 nm. The pH of this solution was pH 7.8 compared to pH 7.9 for the Eskalab methods described above. Finally, 0.02 ml of the s lurry of "Am- monia No. 2" tablet was added and the absorbance re- corded af te r 30 minutes at 37 ° •

Calculations. Values for patient samples were obtained directly from calibration curves performed with each run, similar to those shown in Figure 1. In comparing methods, results obtained on pat ient samples were subjected to least- squares analysis to determine the slope and intercept of the regression line.

Recovery. Aliquots of aqueous ammonia standard solu- tions, prepared from the Hyland nitrogen standard, were added to a pat ient sample (1 to 9 by volume) to examine the percent recovery of ammonia with the Hyland micro- procedure.

RESULTS AND DISCUSSION

Plo ts of the observed changes in absorbance a g a i n s t ammonia concent ra t ion employ ing the methods de-

e " ,,,,," Y , O . 7 6 X ~ 2 9 • r - O . S Z I / / . ."

e •

" : ~ 2"00 I00 H Y L A N D M A C R O - M E T H O D (umol/I]

Fig. 8 - - Compar ison o f H y l a n d Mac- ro- l t~ thod and mod i f i ed Eska lab method.

scr ibed above are shown in Fig . 1. Severa l obse rva t ions f rom th i s f i g u r e inc lude :

1) changes in absorbance wi th ammonia concen t ra t ion a r e f ive fo ld g r e a t e r w i th the H y l a n d K i t (C) ) t h a n with the Eska l ab k i t ( • ) and 2) the Eska lab method may be adap ted to produce much l a r g e r absorbance changes.

A compar ison of the Hyland and Eska lab methods pe r fo rmed on pa t i en t samples accord ing to the m a n u - f a c t u r e r ' s i n s t ruc t ions is shown in Fig . 2. I t is ap- p a r e n t f rom the r eg re s s ion da ta and cor re la t ion coef- f ic ien t the methods a re not in s a t i s f a c t o r y ag reement . Consequently, the sample to Eska l ab r e a g e n t r a t io was modi f ied to yield absorbance r e a d i n g s comparab le to the Hyland method. Compared to the Hyland method, F ig . 3, th is modi f ied Eska lab method did not show an improved agreement .

In developing the enzymic method, Mor idzac et al '~) r e p o r t e d t h a t a p r e - i n c u b a t i o n pe r iod of 25 minu t e s a t room t e m p e r a t u r e was n e c e s s a r y for comple t ion of the s p o n t a n e o u s ox ida t i on of N A D H r e s u l t i n g f rom N A D H - r e q u i r i n g enzymes and s u b s t r a t e s in plasma. Wi th the Eska lab procedure , a number of samples had spur ious absorbance r ead ings even a f t e r 15 minu tes at 37 ° which probab ly con t r ibu ted to the resu l t s in F ig . 3.

Therefore , depro te in iza t ion wi th perchlor ic acid was incorpora ted into the modi f ied Eska l ab method p r i o r to ammonia de t e rmina t i on to remove i n t e r f e r i n g pro- te ins . F ig . 4 shows a plot of the modi f i ed Eska l ab method wi th depro te in iza t ion a g a i n s t the Hy land meth- od. The improved cor re la t ion ( r ~ 0.968) sugges ted the d i s a g r e e m e n t between the methods observed in F igs . 2 and 3 was due to p ro te in i n t e r f e r ence in the Eska l ab procedure . Recently, van Anken and Schip- hors t ") r epor ted t ha t subs t i t u t i on of N A D P H fo r N A D H also may reduce this in te r fe rence .

Ano the r approach to r educ ing the sample size fo r , ammonia de t e rmina t i on was s imply to scale down the Hyland method. A compar ison of "macro" and "micro" Hy land methods, F ig . 5, showed good ag reemen t . I n add i t ion , a recovery exper iment , F ig . 6, revealed the Hy land micro-method quan t i t a t i ve ly recovered all of the ammonia added to a pa t i en t sample. The l i n e a r i t y

194 SAMPSON AND DEMERS

20O

'5 z flO 0

tu

o

X

r • 0.968

,~o ' 26o HYLAND MACRO-METHOD (urnol/I)

Fig. 4 - - Compa~'~son of modified Eskalab method with dep~'oteinization and Hyland macro-method.

300

200

.J

>

0

0

j// 2:5 56 A ,6o

% OF PATIENT SAMPLE

Fig. 7 - - Plot o/ the percent of patient sample after dilw- ~on with water vs the ammonia value measured by the Hyland ~icro-method.

of the micro procedure was verified to 270 /~mol/1 with an elevated patient sample, shown in Fig. 7.

In conclusion, based on comparison studies, either the modified Eskalab method with deproteinization or the Hyland micro-method are acceptable procedures for the routine determination of ammonia with sample

200-

• ~ Y. 1.02 X-2 ~ r-0.986

" Co~" - I I IOO 200

HYLAND MACRO-METHOD (umol/I]

Fig. 5 - - Comparison of Hyland micro- ~,ethod and Hyland macro-method.

5OO

a uJ

~ 2 5 0 •

< • . * 8 6 1

' 260 ' 4bo AMMONIA A D D E D (umol/I)

Fig. 6 - - Plot of ammonia concentra- tion added to a patient sample vs the ammonia value measured by the Hy- land micro-,nethod.

volume of 100 /zl. We have employed the latter proce- dure in our laboratory for approximately one year and found it less tedious to perform than either the enzy- matic method with deproteinization or the unmodified Hyland procedure.

ACKNOWLEDGEMENT

This work was supported in part by a postdoctoral training grant, No. 1F 22 AMO2827-01, NIH, USPHS. The technical assistance of Patti Hicks, ASCP and Richard Walters is acknowledged.

REFEBEN CES

1. Kingsley, G. R. and Tager, H. S., (1970). Standard Methods of Clinical Chemistry, Vol. 6, Academic Press, New York, N.Y., p. 115-126.

2. Henry, R. J., Cannon, D. C., Winkelman, J. W., (1974). Clinical Chemistry Principles and Techniques, 2nd ed., Harper and Row, p. 518.

3. Moridzac, A., Ehrlick, G. E., Seegmiller, J. E., (1965). J. Lab. a~d Clin. Med., 66, 526-531.

4. Van Anken, H. C., Schiphorst, M. E., (1974). Clin. Chem. Acts., 56, 151-157.