Exploring metabolism flexibility in complex organisms through quantitative study of precursor sets for system outputs

Table 8 Effect of long-chain fatty acids (LCFA) oxidization in the triacarboxylic acid (TCA) cycle over model analysis

Main characteristics of models with different ratios of LCFA oxidized in TCA
Model		CO2 prediction		ATP	ATP = 1250 mmol/h/half udder
				optimum	Extreme flux distribution (see Table 9)			Variability of AIO coefficients
Ratio of	Dataset	Predicted	Ratio of		Total number	Nonplausible	Nonplausible	Glucose carbon	Glucose carbon
long-chain FA		CO2	predicted CO2			flux values	AIO	required to	oxidized
oxidated in TCA			and measured CO2					produce lactose
0%	(HB)	1546	Non available	6628	Nonrelevant hypothesis [20–22]
	(CN)	1021	99%	2045	8	6	2	[62.2 ; 72.0]	[17.5 ; 28.9]
	(Ctrl)	1126	121%	3081	8	6	2	[52.8 ; 62.3]	[27.8 ; 39.7]
10%	(HB)	1640	Non available	7395	13	13	0	[47.7 ; 62.4]	[28.8 ; 45.9]
	(CN)	1107	107%	2739	8	6	2	[59.4 ; 72.0]	[17.5 ; 32.0]
	(Ctrl)	1202	129%	3701	8	6	2	[51.6 ; 62.3]	[27.9 ; 41.2]
20%	(HB)	1756	Non available	8336	13	13	0	[46.9 ; 62.4]	[29.0; 47.0]
	(CN)	1214	118%	3607	8	6	2	[57.3 ; 72.0]	[17.5 ; 34.6]
	(Ctrl)	1298	140%	4476	Nonrelevant hypothesis: predicted CO2 is not compatible with measured CO2 [28]
25%	(HB)	1826	Non available	8396	13	13	0	[46.5; 62.4]	[29.0; 47.5]
	(CN)	1279	124%	4128	8	6	2	[56.3 ; 72.0]	[17.6 ; 35.8]
	(Ctrl)	1355	146%	4938	Nonrelevant hypothesis: predicted CO2 is not compatible with measured CO2 [28]
				ATP balance is too high	Extreme distributions are not biologically relevant	For plausible ratios of long-chain FA in TCA,(CN) treatment is characterized by a lower proportion of glucose (on a carbon basis) which is oxidized in CO2, and a larger ratio used for lactose synthesis.

To study the impact of the variability of FA oxidation, a ratio of long-chain FA (10%, 20%, 25%) was introduced in the TCA cycle. For datasets (CN), (Ctrl) and (HB) which were compatible with a given ratio of LCFA oxidation, extreme flux distributions and AIO coefficients variability were studied. Our main conclusions are robust to the introduction of LCFA oxidation (Table 8). Interestingly, as shown in Table 9, the structure of the simplex generated by the (HB) diet is more complicated than the (Ctrl) and (CN) treatments, with 13 vertices for all hypotheses. This is due to the fact that R ₁₅ is highly constrained by measurements, so that this flux cannot vanish when R ₁₄ is optimized or when R ₁₉ is maximized. All the vertices for the (HB)-simplex contradict knowledge-based literature.

ISSN: 1752-0509