Simultaneous analysis of potassium and ammonium ions in the vitreous humour by capillary electrophoresis and their integrated use to infer the post mortem interval (PMI)

Abstract

Post-mortem changes of ions in the body fluids have been proposed as an objective tool for inferring the time of death. In particular, the post-mortem increase of potassium concentrations in the vitreous humour has gained great attention in the literature. On the other hand, ammonium, another ion released in post-mortem processes, has received much less attention, potentially due to unresolved analytical issues using current clinical chemistry methods. This paper presents an application of a new analytical approach based on capillary electrophoresis providing the simultaneous analysis of potassium and ammonium ions in the vitreous humour. In addition, to assess the consistency of the post-mortem increase of ammonium concentrations in the vitreous humour, the determination of this ion in the vitreous humour of the two eyes of the same body at the same post-mortem interval has been verified. Vitreous humour was collected from 33 medico-legal cases where the time of death was known exactly. Prior to analysis, all samples were diluted 1:20 with a 40 μg/mL solution of BaCl₂ (internal standard). In the study of the variability of the ammonium concentration between the two eyes, no statistically significant differences were found, supporting the hypothesis of an even post-mortem increase of the ion concentrations in this particular biological fluid. Significant correlations of potassium and ammonium ions with the post-mortem interval were found, with r² of 0.75 and 0.70, respectively.

Keywords

Ammonium capillary electrophoresis post mortem interval potassium vitreous humour

Introduction

The estimation of time of death remains one of the major challenges in criminal investigations and is often an issue that is debated in the courts. In the past decades, active research has been carried out to identify sound and objective parameters upon which the estimation of the post mortem interval (PMI) could be grounded.^1,2

In fact, PMI estimation is traditionally based on the evaluation of the body’s physical changes occurring in the ‘early post-mortem period’, comprising ‘rigor mortis’, lividities, body cooling and, in some cases, the so-called supravital phenomena. This approach, however, shows important points of weakness, since lividities, rigor mortis and, to some extent, supravital phenomena suffer from high variability and often subjectivity of recording. Also, body cooling is greatly affected by several confounding factors (e.g. environmental temperature and humidity, clothing, posture, body temperature before death, body weight), which substantially limit its reliability in real cases.³ Moreover, the equilibration of the body temperature with the environment is almost completed within the first 20 hours after death, after which the method is useless.

Other, arguably more objective, approaches are based on the study of the time-correlated chemical changes occurring in corpses and are grouped under the collective name of ‘thanatochemistry’.² At present, vitreous humour is by far the preferred biological fluid used in this context, because of its isolation from the environment, ease of collection and protection from contamination by external agents. Numerous compounds have been investigated in the vitreous humour,⁴ but so far the most studied is potassium. Its concentration in the vitreous humour increases after death because of passive release from the intracellular compartment into the vitreous humour (substantially an extracellular fluid), until an equilibrium is reached between the two compartments. From the early 1960s until now, many studies have proposed the potassium increase as a tool to infer the PMI.^5,6 However, discrepancies in research have been reported concerning the correlation potassium-PMI with regard to the intercept and the slope, which have limited the practical application of the method. These discrepancies can be attributed to several factors, such as different causes of death, different storage conditions of the bodies and different modes of sample collection, but, above all, to the different analytical techniques employed, most of which were not specifically validated for post-mortem analysis.⁶ This point applies particularly to the forensic samples, the composition of which is typically non-standardised and possibly affected by numerous uncontrolled variables and anomalies. Moreover, only a minority of authors have used separation methods, for example ion chromatography (IC) and capillary electrophoresis (CE), specifically developed for this purpose.^8–11 In fact, the majority of authors used automated non-separation clinical chemistry techniques for the analysis of serum and urine (mostly based on ion selective electrodes), which are inherently prone to be interfered by the matrix composition.⁷

On the other hand, ammonium is another ion that typically increases post-mortem as a product of protein deamination. Surprisingly, it was neglected by most of authors for quite obscure reasons, but probably related to well-known analytical problems in clinical chemistry and laboratory medicine.^12–18 To the best of our knowledge, a single paper did appear in 1978 reporting an ammonium increase in the vitreous humour in correlation with the post mortem interval. In this article, a colorimetric method based on the Nessler’s reagent was used, without any separation step.¹⁶ Only recently, a CE method for the determination of ammonium in the vitreous humour with indirect ultraviolet (UV) detection has been reported, showing a potential for use in this context.¹⁷ This simple CE separation method had previously proved suitable for the separation and quantitative determination of potassium ions in the vitreous humour.⁹ In addition, two more methodologies developed by Musile et al., a gas diffusion device and a microfluidic paper-based device, were recently developed in our research group.^18,19 The aim of both is the determination of NH₄⁺ in vitreous humour. Both of them are portable and provide in situ qualitative results which will need to be confirmed by CE. The more powerful of the two is the microfluidic paper-based device because it has a faster analysis time, multiple samples can be measured simultaneously and the specificity is higher since the Nessler’s reagent is used (which is selective to NH₄⁺).

The present work was designed to verify, in real cases, the applicability and usefulness of the simultaneous analysis of these two ions using CE to achieve a more reliable estimation of the post-mortem interval and consequently of the time of death. The present method validation also includes a specific study aimed at verifying the consistency of the post-mortem increase of ammonium concentrations in the vitreous humour by comparing the results from the two eyes at the same time after death.

Materials and methods

Standards and chemicals

All chemicals were of analytical-reagent grade. Standard solutions of ammonium (NH₄⁺), potassium (K⁺) and barium (Ba²⁺) were prepared from AnalaR salts (NH₄Cl, KCl and BaCl₂) (Merck, Darmstadt, Germany). 18-crown-6 ether (99% pure) and α-hydroxybutyric acid (HIBA) (99% pure) were obtained from Aldrich (Milan, Italy); Imidazole (99% pure) was obtained from Sigma (St Louis, MO, USA). The electrophoretic buffer pH was adjusted to the desired pH with glacial acetic acid (Merck, Germany). MilliQ water (18.2 MΩ) was used throughout all the experiments.

Instrumentation

All experiments were performed using a P/ACE MDQ Capillary Electrophoresis System (Beckman, Fullerton, CA, USA) equipped with a UV filter detector set at 214 nm wavelength. In all the experiments, untreated fused-silica capillaries (75 μm I.D., 50 cm effective length; Beckman) were used with a detection window of 200 × 100 μm. The capillary was thermostated at 25°C with a perfluorinated coolant flowing in the capillary cartridge. Beckman P/ACE Station (version 8.0) software was used for instrument control, data acquisition and processing.

Electrophoresis conditions

Electrophoresis separations were performed using a running buffer composed of 5 mM imidazole, 6 mM HIBA and 5 mM 18-crown-6 ether adjusted to pH 4.5. Constant voltage runs were performed in all experiments by applying a field of 500 V/cm at 25°C with a resulting current of about 18 μA. In order to overcome the lack of optical absorbance of ammonium, ‘indirect UV detection’ was applied, using a separation buffer absorbing in the UV region. The used detection wavelength was 214nm, with automatic inversion of the signal in order to record peaks mimicking ordinary UV detection. The analytes were injected at the anodic end of the capillary at 0.5 psi for 10 seconds. New naked silica capillaries were washed with NaOH 1 mol/L (30 minutes) and then with deionised water for 25 minutes. Every day the capillary was rinsed with a solution containing NaOH 1 mol/L (20 minutes) followed by deionised water (15 minutes) and finally conditioned with the running buffer (10 minutes). Between consecutive runs the capillary was washed with water (3 minutes) and then with the running buffer (2 minutes). Water blanks were routinely checked for contamination by trace amounts of cations.

Sample collection and preparation

Vitreous humour samples (100–200 μL) were collected by puncture of the sclera at the lateral cantus of the eye from 33 forensic cases of violent or sudden deaths, in which the time of death was exactly known (age range 19–95 years). Cases of traumas involving the head and particularly the eyes were excluded. On the basis of previous research by Bortolotti et al. reporting that in primary acute respiratory failure the vitreous humour potassium concentration is increased,²⁰ the cases of death for opioid overdose were also excluded.

Before analysis, the samples were stored at −24°C for no longer than a week. For the analysis of both ions, NH₄⁺ and K⁺, the vitreous humour was simply diluted 1:20 with an aqueous solution of barium (40 μg/mL), which was used as the internal standard (IS).

Results and discussion

Calibration curves of ammonium and potassium

The details of the validation of the analytical methods for the determination of NH₄⁺ and K⁺ in vitreous humour used in the present work were reported in previous papers.^17,21

Since the present work was aimed at the validation of the simultaneous determination of the two ions in post mortem cases, in a first step, calibration curves for ammonium and potassium were produced on the basis 33 medico-legal autopsies; the results are displayed in Figures 1 and 2, respectively, showing a statistically significant relationship between both ions and PMI. The correlation is better described with polynomial equations as follows:

Ammonium : y = 2×1 0^{- 6} x^{2} + 0.0 127x + 0. 1461

Potassium : y = - 0.000 {5x}^{2} + 0. 20 18x+6 . 173

Figure 1.

Second degree polynomial correlation between PMI (hours) and vitreous NH₄⁺ concentration (mM) determined in 33 forensic cases with known time of death. Equation y = 2×10⁻⁶x² + 0.0127x + 0.1461; r² = 0.70 in which x = PMI (hours); y = NH₄⁺ concentration (mM).

Figure 2.

Correlation between PMI (hours) and vitreous K⁺ concentration (mM) determined in 33 forensic cases with known time of death. Equation y = −0.0005x² + 0.2018x + 6.173; r² = 0.75 in which x = PMI (hours); y = K⁺ concentration (mM).

The correlation coefficient for vitreous humour ammonium vs. PMI is r² = 0.70 and that for potassium is r² = 0.75.

An integrated use of these two parameters by using the average of potassium and ammonium concentrations has also been tested (see Figure 3). In order to balance the weight of the two parameters in the calculation, the logarithm of the potassium concentration was used. This approach led to a small increase of the correlation coefficient (r² = 0.74).

Figure 3.

Correlation between PMI (hours) and combined vitreous NH₄⁺ and K⁺ concentration (mM) determined in 33 forensic cases with known time of death. Equation y = −0.00001x² + 0.0106x + 0.4799; r² = 0.74 in which x = PMI (hours); y = NH₄⁺ and K⁺ concentration combined (mM).

In order to calculate the experimental error in a practical use of the method, each case was treated, one by one, as ‘unknown’ and the theoretical PMI was calculated after having excluded its ion concentrations from the calibration curve. The theoretical PMIs were compared to the observed ones and the differences recorded (Tables 1 to 3). The average absolute error of estimation for potassium and ammonium was 17.8 hours (SD = 15.3) and 20.2 hours (SD = 15.6), respectively.

Table 1.

Estimation of error for potassium. The coefficients of the curve are denoted by a,b,c, (y = ax² + bx + c). Estimated PMI is obtained as the solution to the quadratic equation.

VH sample	[K⁺] (mM)	PMI (hours)	a	b	c	Estimated PMI (hours)	Absolute error (hours)	Relative error	Error as % PMI
1	15.28	77.1	−0.0006	0.2130	5.9396	51.3	−25.8	−0.34	−33.5
2	14.03	47	−0.0005	0.2023	6.1764	43.5	−3.5	−0.08	−7.5
3	12.30	21.5	−0.0005	0.2078	5.8904	33.6	12.1	0.56	56.1
4	24.47	143	−0.0005	0.2030	6.1570	135.2	−7.8	−0.05	−5.4
5	11.94	46	−0.0006	0.2044	6.2006	30.9	−15.1	−0.33	−32.9
6	22.01	80.45	−0.0005	0.1904	6.4179	119.2	38.7	0.48	48.1
7	4.67	17.45	−0.0005	0.1830	6.9891	−12.3	−29.7	−1.70	−170.3
8	17.43	123.45	−0.0006	0.2093	5.9239	68.4	−55.1	−0.45	−44.6
9	8.62	32	−0.0005	0.1986	6.4336	11.3	−20.7	−0.65	−64.6
10	16.47	55	−0.0005	0.2000	6.1979	60.5	5.5	0.10	10.0
11	14.95	43	−0.0005	0.2010	6.1465	50.0	7.0	0.16	16.4
12	27.23	168.5	−0.0009	0.2412	5.3116	162.5	−6.0	−0.04	−3.6
13	7.27	16.1	−0.0005	0.1930	6.5324	3.9	−12.2	−0.76	−76.0
14	13.56	47.2	−0.0006	0.2029	6.1795	41.5	−5.7	−0.12	−12.2
15	21.18	100.3	−0.0005	0.2008	6.1971	99.0	−1.3	−0.01	−1.3
16	8.04	29	−0.0005	0.1967	6.4968	8.0	−21.0	−0.72	−72.4
17	18.98	96.45	−0.0006	0.2072	6.0412	81.8	−14.6	−0.15	−15.1
18	17.43	52.3	−0.0005	0.1978	6.1987	68.7	16.4	0.31	31.4
19	23.66	99.45	−0.0005	0.1922	6.4117	142.8	43.3	0.44	43.6
20	16.43	98.45	−0.0006	0.2165	5.8112	58.5	−39.9	−0.41	−40.5
21	20.11	99.16	−0.0006	0.2041	6.1162	95.2	−3.9	−0.04	−4.0
22	23.73	143.4	−0.0005	0.2016	6.1756	127.2	−16.2	−0.11	−11.3
23	10.75	25	−0.0005	0.2015	6.1875	24.1	−0.9	−0.04	−3.7
24	14.08	22.24	−0.0006	0.2120	5.6808	45.5	23.2	1.04	104.5
25	10.66	23.35	−0.0005	0.2020	6.1643	23.6	0.3	0.01	1.1
26	9.03	17.55	−0.0005	0.1998	6.2599	14.4	−3.2	−0.18	−18.1
27	18.39	47.4	−0.0005	0.1968	6.1494	77.4	30.0	0.63	63.4
28	15.86	47.4	−0.0005	0.2001	6.1652	56.4	9.0	0.19	19.0
29	13.67	50.15	−0.0006	0.2037	6.1687	42.0	−8.1	−0.16	−16.2
30	20.58	48.45	−0.0005	0.1935	6.1602	100.7	52.3	1.08	107.9
31	5.74	6.55	−0.0005	0.1893	6.6474	−4.7	−11.3	−1.72	−172.3
32	18.51	75.5	−0.0005	0.2010	6.1899	75.5	0.0	0.00	−0.1
33	16.57	27.35	−0.0006	0.2110	5.6377	63.2	35.8	1.31	130.9

Table 2.

Estimation of error for ammonium. The coefficients of the curve are denoted by a,b,c, (y = ax² + bx + c). Estimated PMI is obtained as the solution to the quadratic equation.

VH sample	[NH₄⁺] (mM)	PMI (hours)	a	b	c	Estimated PMI (hours)	Absolute error (hours)	Relative error	Error as % PMI
1	0.99	77.1	−5.00E-07	0.0132	0.1356	65.2	−11.88	−0.15	−15.41
2	1.01	47	5.00E-06	0.0123	0.1440	68.7	21.74	0.46	46.26
3	0.51	21.5	1.00E-06	0.0129	0.1340	28.8	7.35	0.34	34.16
4	1.36	143	2.00E-05	0.0115	0.1617	90.3	−52.71	−0.37	−36.86
5	0.31	46	1.00E-06	0.0131	0.1510	12.1	−33.87	−0.74	−73.64
6	1.49	80.45	9.00E-06	0.0115	0.1702	106.1	25.66	0.32	31.90
7	0.31	17.45	4.00E-06	0.0124	0.1560	12.4	−5.08	−0.29	−29.11
8	1.47	123.45	2.00E-06	0.0131	0.1330	100.2	−23.24	−0.19	−18.83
9	0.39	32	3.00E-06	0.0125	0.1590	18.0	−13.99	−0.44	−43.73
10	0.97	55	4.00E-06	0.0124	0.1487	64.7	9.74	0.18	17.72
11	0.38	43	3.00E−07	0.0129	0.1537	17.6	−25.41	−0.59	−59.10
12	2.29	168.5	1.00E-05	0.0117	0.1667	159.5	−9.01	−0.05	−5.35
13	0.31	16.1	3.00E-06	0.0125	0.1535	12.5	−3.62	−0.22	−22.47
14	0.60	47.2	1.00E-06	0.0129	0.1472	35.3	−11.87	−0.25	−25.15
15	1.49	100.3	3.00E-06	0.0125	0.1502	104.2	3.89	0.04	3.88
16	0.35	29	3.00E-06	0.0124	0.1614	15.2	−13.85	−0.48	−47.74
17	1.38	96.45	2.00E-06	0.0127	0.1453	95.8	−0.67	−0.01	−0.70
18	1.42	52.3	1.00E-05	0.0116	0.1532	100.5	48.20	0.92	92.16
19	2.14	99.45	1.00E-05	0.0103	0.2057	162.2	62.79	0.63	63.14
20	0.82	98.45	−8.00E-06	0.0147	0.0958	50.7	−47.79	−0.49	−48.54
21	1.06	99.16	−4.00E-06	0.0139	0.1154	69.3	−29.82	−0.30	−30.07
22	2.98	143.4	2.00E-05	0.0145	0.1207	161.3	17.90	0.12	12.49
23	0.38	25	3.00E-06	0.0125	0.1559	17.9	−7.15	−0.29	−28.59
24	0.31	22.24	4.00E-06	0.0123	0.1621	12.0	−10.26	−0.46	−46.14
25	0.31	23.35	4.00E-06	0.0123	0.1630	11.9	−11.44	−0.49	−49.01
26	0.31	17.55	4.00E-06	0.0124	0.1561	12.4	−5.19	−0.30	−29.56
27	0.99	47.4	5.00E-06	0.0124	0.1447	66.4	18.99	0.40	40.07
28	0.59	47.4	9.00E-07	0.0129	0.1472	34.2	−13.16	−0.28	−27.76
29	0.56	50.15	1.00E-07	0.0130	0.1453	31.9	−18.26	−0.36	−36.41
30	1.23	48.45	7.00E-06	0.0120	0.1451	86.1	37.64	0.78	77.68
31	0.31	6.55	−6.00E-07	0.0133	0.1241	14.0	7.44	1.14	113.53
32	0.91	75.5	−2.00E-06	0.0134	0.1315	58.6	−16.89	−0.22	−22.37
33	1.02	27.35	−1.00E-06	0.0136	0.0929	68.5	41.16	1.51	150.51

Table 3.

Estimation of error for for the combined used of ammonium and potassium. The coefficients of the curve are denoted by a,b,c, (y = ax² + bx + c). Estimated PMI is obtained as the solution to the quadratic equation.

VH sample	[NH₄⁺] + log[K⁺] (mM)	PMI (hours)	a	b	c	Estimated PMI (hours)	Absolute error (hours)	Relative error	Error as % PMI
1	1.09	77.1	−2.00E-06	0.0092	0.5089	64.0	−13.13	−0.17	−17.03
2	1.08	47	8.00E-07	0.0087	0.5146	64.6	17.61	0.37	37.46
3	0.80	21.5	−2.00E-06	0.0091	0.5038	32.6	11.11	0.52	51.69
4	1.38	143	4.00E-06	0.0087	0.5152	94.8	−48.21	−0.34	−33.71
5	0.69	46	−2.00E-06	0.0091	0.5200	19.1	−26.85	−0.58	−58.38
6	1.42	80.45	3.00E-06	0.0082	0.5308	104.1	23.68	0.29	29.43
7	0.49	17.45	3.00E-06	0.0082	0.5460	−6.9	−24.34	−1.39	−139.47
8	1.35	123.45	−2.00E-06	0.0093	0.5017	93.5	−29.97	−0.24	−24.28
9	0.66	32	2.00E-08	0.0087	0.5271	15.3	−16.70	−0.52	−52.18
10	1.09	55	8.00E-07	0.0087	0.5180	65.6	10.64	0.19	19.35
11	0.78	43	−1.00E-06	0.0089	0.5197	29.1	−13.92	−0.32	−32.37
12	1.86	168.5	8.00E-06	0.0078	0.5390	147.3	−21.24	−0.13	−12.61
13	0.59	16.1	1.00E-06	0.0086	0.5291	6.6	−9.52	−0.59	−59.10
14	0.87	47.2	−1.00E-06	0.0089	0.5170	39.4	−7.80	−0.17	−16.52
15	1.41	100.3	−3.00E-07	0.0088	0.5171	101.6	1.29	0.01	1.28
16	0.63	68	−8.00E-05	0.0087	0.5146	15.1	−52.92	−0.78	−77.82
17	1.33	96.45	−1.00E-06	0.0090	0.5190	90.9	−5.51	−0.06	−5.72
18	1.33	52.3	3.00E-06	0.0083	0.5132	95.2	42.91	0.82	82.05
19	1.76	99.45	6.00E-06	0.0076	0.5477	143.0	43.53	0.44	43.77
20	1.02	98.45	−7.00E-06	0.0101	0.4849	54.8	−43.60	−0.44	−44.29
21	1.18	99.16	−4.00E-06	0.0096	0.4985	73.4	−25.75	−0.26	−25.97
22	2.18	143.4	−5.00E-06	0.0091	0.5170	205.7	62.35	0.43	43.48
23	0.71	25	−1.00E-07	0.0088	0.5199	21.1	−3.88	−0.16	−15.52
24	0.73	22.24	−6.00E-07	0.0089	0.5142	24.2	1.97	0.09	8.85
25	0.67	23.35	2.00E-07	0.0087	0.5231	16.7	−6.61	−0.28	−28.31
26	0.63	17.55	3.00E-07	0.0087	0.5227	12.7	−4.90	−0.28	−27.89
27	1.13	47.4	1.00E-06	0.0086	0.5144	70.7	23.29	0.49	49.13
28	0.90	47.4	−7.00E-07	0.0089	0.5167	42.7	−4.73	−0.10	−9.99
29	0.85	50.15	−1.00E-06	0.0090	0.5163	37.0	−13.16	−0.26	−26.23
30	1.27	48.45	2.00E-06	0.0084	0.5144	88.3	39.85	0.82	82.25
31	0.53	6.55	9.00E-07	0.0086	0.5268	0.9	−5.66	−0.86	−86.41
32	1.09	75.5	−2.00E-06	0.0092	0.5101	63.8	−11.72	−0.16	−15.53
33	1.12	27.35	−3.00E-06	0.0095	0.4797	68.9	41.51	1.52	151.78

When the concentrations of both ions were combined, the average absolute error was 21.5 hours (SD = 16.5). The average relative error, expressed as percentage of the PMI (%PMI), was 44%, 42% and 43% for potassium, ammonium and the combined concentration of both. It is patent that the relative errors decrease steadily with the increasing of the PMIs, from about 200% in the first hours down to low percentages for PMIs above 100 hours (Figure 4).

Figure 4.

Estimation error (expressed as %PMI) with the PMI (hours) for potassium, ammonium and the sum of the concentrations of both ions.

Ammonium concentration differences in the two eyes

The statistical difference between the ammonium concentrations in the two eyes of the same subject was evaluated in 20 forensic autopsies. The mean concentrations of NH₄⁺ ranged from 0.07 to 6.45 mM with PMIs varying from 31 up to 89 hours. The differences in concentration between the two eyes were evaluated using a paired student’s t-test (α = 0.01). The findings suggest that no statistically significant differences exist (p value for two tails = 0.87). In addition, a good correlation was found between the two eyes, expressed by r² = 0.94).

Discussion and conclusion

The present study confirms the time-dependent increase of the concentration of potassium in the vitreous humour, but, more importantly, also shows that ammonium has a potential as a marker of time elapsed since death, confirming recent preliminary findings.¹⁷

It is worth mentioning that, even with a very fast sample pre-treatment (just dilution 1:20), the described CE method provides very clean electropherograms, as shown in Figures 5 and 6 representing two cases with PMIs of 32 and 99 hours, respectively. The comparative study of the correlation of the two parameters with the PMI and of the possibility of their use to infer the time of death in real forensic cases shows a superiority of potassium in comparison with ammonium. However, in our opinion ammonium can complement potassium for the estimation of PMI, particularly when potassium concentrations in the vitreous humour could have been affected by confounding factors, for example ocular traumas or drug overdoses. In fact, because of the different mechanisms at the basis of the post-mortem increases of the two ions, the carrying out of their determination in parallel can provide mutual corroboration in terms of reliability of the estimated PMIs.

Figure 5.

Typical electropherogram of human vitreous humour (NH₄⁺ = 0.39; K⁺ = 8.62; PMI = 32 hours)

Figure 6.

Typical electropherogram of human vitreous humour (NH₄⁺ = 1.18; K⁺ = 20.81; PMI = 99 hours)

In regards to the error related to the determination, our data show that as the PMIs increase so do the relative accuracy of its estimation, as depicted in Figure 4. This confirms previous findings reporting the variability of PMI estimation for PMIs below 26 hours.²²

The main limitation of this study is the still small number of the cases investigated, justified by obvious difficulties in the selection of traumatic or sudden deaths with known PMI. A second limitation is related to the error of the determination. As it can be seen in Figure 4, at early PMIs (up to 60 hours) the error expressed as the percentage of PMI is very high. However, as outlined in the introduction, for the determination of short PMIs there are other methodologies based on the evaluation of the body’s physical changes occurring in the ‘early post-mortem period’.

The present study was also aimed at verifying the consistency of ion concentrations between the two eyes. A similar study was carried out by our group on potassium concentrations and reported no significant difference.⁷ The present data show that there is also no substantial difference in the concentrations of ammonium between the two eyes. This evidence supports the hypothesis that there is an even progression of the ammonium ion production in this body compartment.

In conclusion, our findings support a potential usefulness of ammonium determination in the vitreous humour, in addition to potassium, as an objective biochemical parameter to infer the PMI, and the suitability of capillary electrophoresis to perform rapid simultaneous determinations.

Future research should therefore concentrate on the acquisition of more vitreous humour samples and the further analysis to build a more reliable statistical method that can be used in the analysis of real unknown samples.

Footnotes

Competing interests

The funding organisation(s) played no role in the study design; in the collection, analysis and interpretation of data; in the writing of the report; or in the decision to submit the report for publication.

Contributorship statement

Palacio and Gottardo performed the analysis, processed the experimental data, drafted the manuscript and designed the figures. Bortolotti and Musile aided in interpreting the results and worked on the manuscript. Cirielli was involved in the sample collection. Yvane helped with the data and revision of the manuscript. Tagliaro and Bortolotti were involved in planning and supervised the work. All authors discussed the results and commented on the manuscript.

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship and/or publication of this article.

Ethical approval

Collection of samples from cases was carried out in a forensic context. Consequently no ethical approval was needed in our jurisdiction. In addition all data were anonymised.

Funding

The authors received no financial support for the research, authorship and/or publication of this article.

ORCID iDs

Covadonga Palacio

Vito Cirielli

Giacomo Musile

Franco Tagliaro

References

Madea

Methods for determining time of death.

Forensic Sci Med Pathol 2016; 12: 451–485.

Madea

Estimation of the time since death. Boca Raton: CRC Press, 2015.

Hubig

Muggenthaler

Sinicina

, et al. Temperature based forensic death time estimation: the standard model in experimental test. Leg Med 2015; 17(5): 381–387.

Thierauf

Musshoff

Madea

Post-mortem biochemical investigations of vitreous humor.

Forensic Sci Int 2009; 192: 78–82.

Coe

JI.

Vitreous potassium as a measure of the postmortem interval: an historical review and critical evaluation.

Forensic Sci Int 1989; 42: 201–213.

Madea

Is there recent progress in the estimation of the postmortem interval by means of thanatochemistry?

Forensic Sci Int 2005; 151: 139–49.

Tagliaro

Bortolotti

Manetto

, et al. Potassium concentration differences in the vitreous humour from the two eyes revisited by microanalysis with capillary electrophoresis. J Chromatogr A 2001; 924: 493–498.

Zhou

Zhang

, et al.

The determination of potassium concentration in vitreous humor by low pressure ion chromatography and its application in the estimation of postmortem interval.

J Chromatogr B Anal Technol Biomed Life Sci 2007; 8522: 278–281.

Nikolac

Omazic

Simundic

AM.

The evidence based practice for optimal sample quality for ammonia measurement.

Clin Biochem 2014; 47: 991–995.

10.

Schuff-Werner

Steiner

Preanalytical ammonia generation: a race with time but not with temperature.

J Clin Pathol 2015; 68: 757–758.

11.

Fonseca-Wollheim

Deamidation of glutamine by increased plasma y-glutamyltransferase is a source of rapid pmmonia formation in blood and plasma specimens. Clin Chem 1990; 36: 1479–1482.

12.

Dukic

Simundic

Short-term and long-term storage stability of heparin plasma ammonia.

J Clin Pathol 2015; 68: 288–291.

13.

Fonseca-Wollheim

Preanalytical increase of ammonia in blood specimens from healthy subjects. Clin Chem 1990; 36: 1483–1487.

14.

Wilkinson

Smeeton

Watt

PW.

Progress in neurobiology ammonia metabolism, the brain and fatigue: revisiting the link.

Prog Neurobiol 2010; 91: 200–219.

15.

Howanitz

Howanilz

Skrodzki

, et al. Influences of specimen processing and storage conditions on results for plasma ammonia. Clin Chem 1984; 30: 906–908.

16.

Van den Oever

Post-mortem vitreous ammonium concentrations in estimating the time of death.

Zeitschrift für Rechtsmedizin 1978; 80: 259–263.

17.

Gottardo

Palacio

Shestakova

, et al. A new method for the determination of ammonium in the vitreous humour based on capillary electrophoresis and its preliminary application in thanatochemistry. Clin Chem Lab Med 2019; 57(4) :504–509

18.

Musile

Gottardo

Palacio

, et al. Development of a low cost gas diffusion device for ammonia detection in the vitreous humor and its preliminary application for estimation of the time since death. Forensic Sci Int 2019; 295: 150–156.

19.

Musile

Agard

De Palo

, et al.

Thanatochemistry at the crime scene: a microfluidic paper-based device for ammonium analysis in the vitreous humor.

Anal Chim Acta 2019; 1083: 150–156.

20.

Bortolotti

Gottardo

Trettene

, et al. Increased postmortem concentrations of K+in the vitreous humour in heroin overdose deaths. Annales de Toxicologie Analytique 2002; 14: 306.

21.

Tagliaro

Manetto

Cittadini

, et al. Capillary zone electrophoresis of potassium in human vitreous humour: Validation of a new method. J Chromatogr B Biomed Sci Appl 1999; 733: 273–279.

22.

Pekka Sauco and Bernard Knight. Knight’s Forensic Pathology. CRC press, 3rd ed. 2004.