Skip to contents

Wuzhi-capsule schizandrol lignans as CYP3A4/CYP3A5 inhibitors (He 2022)

Source: vignettes/articles/He_2022_tacrolimus_wuzhi_schizandrol_cyp3a.Rmd
He_2022_tacrolimus_wuzhi_schizandrol_cyp3a.Rmd

Model and source

  • Citation: He Q, Bu F, Wang Q, Li M, Lin J, Tang Z, Mak WY, Zhuang X, Zhu X, Lin HS, Xiang X. Examination of the Impact of CYP3A4/5 on Drug-Drug Interaction between Schizandrol A/Schizandrol B and Tacrolimus (FK-506): A Physiologically Based Pharmacokinetic Modeling Approach. Int J Mol Sci. 2022;23(9):4485. doi:10.3390/ijms23094485. PMCID: PMC9103789. Reversible inhibition constants (Ki): Results 2.2 and Figure 2C,D (SZB on CYP3A4 Ki = 2.18 uM, on CYP3A5 Ki = 2.03 uM); pooled CYP3A Ki = 5.82 uM (Figure 2B). NOTE: the Discussion text swaps the CYP3A4 and CYP3A5 Ki values (it prints CYP3A4 Ki = 2.03, CYP3A5 Ki = 2.18); Results 2.2 and the Figure 2C,D annotations are used here (CYP3A4 = 2.18, CYP3A5 = 2.03). Time-dependent inactivation constants: Results 2.3, Discussion, and the figure annotations (CYP3A4 kinact = 0.37 /min, KI = 0.69 uM, Figure 5B; CYP3A5 kinact = 0.009 /min, KI = 0.5 uM, Figure 5D; pooled CYP3A kinact = 0.044 /min, KI = 0.43 uM, Figure 3D). IC50 shift 21.39 (IC50 11.98 uM no preincubation, 0.56 uM after NADPH preincubation): Results 2.1. Michaelis-Menten inactivation form kobs = kinact*I/(KI+I): Materials and Methods 4.4 and Equation (3).
  • Article: https://doi.org/10.3390/ijms23094485
  • PubMed Central open-access copy: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9103789/

He 2022 co-administer the Wuzhi capsule (a standardised ethanolic extract of Schisandra sphenanthera) with tacrolimus (FK-506), a combination used in Chinese transplant practice because the extract’s lignans inhibit CYP3A and raise tacrolimus exposure, cutting the required tacrolimus dose. An earlier paper by the same group (He 2021, PMC7997453) measured the two most abundant lignans, schisantherin A and schisandrin A. This paper measures the other two constituents, schizandrol A (SZA) and schizandrol B (SZB): their reversible (RI) and time-dependent (TDI) inhibition of CYP3A4 and CYP3A5 separately, using CYP3A5-genotyped human liver microsomes with CYP3cide to isolate CYP3A5.

This paper contributes two model files, one per lignan:

mod_sza <- rxode2::rxode2(readModelDb("He_2022_schizandrolA_cyp3a"))
mod_szb <- rxode2::rxode2(readModelDb("He_2022_schizandrolB_cyp3a"))

# Fixed and estimated ini() values are not returned as columns by rxSolve(), so
# read them straight off the packaged model when a check needs one.
ini_val <- function(mod, nm) {
  stopifnot(nm %in% mod$iniDf$name)
  mod$iniDf$est[mod$iniDf$name == nm]
}

Scope: what this paper does and does not contribute

He 2022 has two halves, and only one of them is reproducible.

The first half is the authors’ own bench work: reversible-inhibition (RI) and time-dependent-inhibition (TDI) assays that measure the inhibitory potency of SZA and SZB on CYP3A4 and CYP3A5, each isoform in its own microsomal system, plus a pooled-HLM CYP3A screen. Every fitted constant is printed in the Results and Discussion text and annotated on Figures 2-5. That is the layer this vignette validates and that the two model files carry.

The second half is a Simcyp (version 16) whole-body physiologically based model that integrates those constants (together with the physicochemical and disposition parameters of Table 3) to predict the fold-increase in tacrolimus AUC in CYP3A5 expressers and non-expressers (Table 2). That layer is a vendor platform model and is not extracted: the whole-body ODE system is Simcyp’s rather than the authors’, the tissue-partition (Kp) values come from a built-in method, the organ volumes/flows and enzyme abundances are Simcyp database outputs, and no project file is deposited. Reconstructing it would require substituting physiology from outside the source. The section “Published results not reproduced here” records what that half produced.

A load-bearing consequence for the in vitro layer: the two lignans and the two isoforms are inhibited by different combinations of the two mechanisms, and each model file records exactly the combination the paper fitted.

Isoform-resolved inhibition mechanisms fitted by He 2022 (Results 2.2, 2.3; Figures 2, 4, 5). A pooled-HLM CYP3A screen (Figures 2B, 3) is also carried in each file.
Lignan Isoform Reversible (Ki, uM) Time-dependent (kinact /min, KI uM)
Schizandrol A (SZA) CYP3A4 little (not fitted) 0.024, 15.38
Schizandrol A (SZA) CYP3A5 little (not fitted) none (Figure 4C)
Schizandrol B (SZB) CYP3A4 2.18 0.37, 0.69
Schizandrol B (SZB) CYP3A5 2.03 0.009, 0.5

In vitro systems

In vitro systems of He 2022 (Methods 4.3, 4.4).
Assay CYP3A4 system CYP3A5 system Pooled CYP3A Probe
Reversible inhibition (RI) CYP3A53/3 HLM, 0.2 mg/mL CYP3A51/3 HLM + 1.2 uM CYP3cide, 0.2 mg/mL Pooled HLM (22 donors), 0.2 mg/mL Tacrolimus disappearance (Dixon plot)
Time-dependent inhibition (TDI) CYP3A53/3 HLM, 0.5 mg/mL CYP3A51/3 HLM + CYP3cide, 0.5 mg/mL Pooled HLM (22 donors), 0.5 mg/mL Testosterone 6-beta-hydroxylation (double-reciprocal plot)

CYP3cide (PF-04981517) is a selective mechanism-based inactivator of CYP3A4; it is used to knock out CYP3A4 so that the residual testosterone or tacrolimus turnover in the CYP3A51/3 microsomes is attributable to CYP3A5.

Source trace

Every value in the two model files, with the location it came from.

Source trace for every ini() value and every model equation.
Model Quantity Value Source location
He_2022_schizandrolA_cyp3a ki_inact_3a4 15.38 uM Figure 4B annotation; Results 2.3
He_2022_schizandrolA_cyp3a kinact_3a4 0.024 /min Figure 4B annotation; Results 2.3
He_2022_schizandrolA_cyp3a ki_inact_3a (pooled) 15.625 uM Figure 3B annotation; Results 2.3
He_2022_schizandrolA_cyp3a kinact_3a (pooled) 0.029 /min Figure 3B annotation; Results 2.3
He_2022_schizandrolA_cyp3a (CYP3A5 TDI) none Figure 4C (no inactivation observed)
He_2022_schizandrolA_cyp3a (any RI) little Figure 2A (no significant reversible inhibition)
He_2022_schizandrolB_cyp3a ki_3a4 2.18 uM Figure 2C annotation; Results 2.2
He_2022_schizandrolB_cyp3a ki_3a5 2.03 uM Figure 2D annotation; Results 2.2
He_2022_schizandrolB_cyp3a ki_3a (pooled) 5.82 uM Figure 2B annotation; Results 2.2
He_2022_schizandrolB_cyp3a ki_inact_3a4 0.69 uM Figure 5B annotation; Results 2.3
He_2022_schizandrolB_cyp3a kinact_3a4 0.37 /min Figure 5B annotation; Results 2.3
He_2022_schizandrolB_cyp3a ki_inact_3a5 0.5 uM Figure 5D annotation; Results 2.3
He_2022_schizandrolB_cyp3a kinact_3a5 0.009 /min Figure 5D annotation; Results 2.3
He_2022_schizandrolB_cyp3a ki_inact_3a (pooled) 0.43 uM Figure 3D annotation; Results 2.3
He_2022_schizandrolB_cyp3a kinact_3a (pooled) 0.044 /min Figure 3D annotation; Results 2.3
both addSd (fixed 0) 0 Not reported anywhere; see Assumptions
Inactivation-rate form kobs = kinact*I/(KI+I) - Methods 4.4.3, Equation (3)
Inactivation ODE d/dt(E) = -kobs*E - log-linear decay of Figures 3A, 4A, 5A, 5C
Reversible form 1/(1 + I/Ki) - Competitive inhibition, Dixon plots (Figure 2)

The fitted constants live in the annotations of Figures 2-5 and are repeated in the Results and Discussion text; they are the authors’ own reported estimates, not digitised from the plotted points. One transcription caveat: the Discussion paragraph swaps the reversible CYP3A4 and CYP3A5 Ki values for SZB (it prints CYP3A4 Ki = 2.03, CYP3A5 Ki = 2.18), whereas Results 2.2 and the Figure 2C/2D annotations give CYP3A4 Ki = 2.18 and CYP3A5 Ki = 2.03. The figure/Results order is used here.

Structural identities of the inactivation model

The time-dependent inactivation is a first-order decay whose rate constant kobs is a hyperbolic (Michaelis-Menten) function of the inhibitor concentration. These identities are exact by construction and pin the transcription of kinact and KI.

kobs <- function(mod, suffix, I) {
  kinact <- ini_val(mod, paste0("kinact_", suffix))
  KI     <- ini_val(mod, paste0("ki_inact_", suffix))
  kinact * I / (KI + I)
}

inactivation_systems <- list(
  list(mod_sza, "3a4"), list(mod_sza, "3a"),
  list(mod_szb, "3a4"), list(mod_szb, "3a5"), list(mod_szb, "3a")
)

# 1. kobs(0) = 0 exactly (no inhibitor => no inactivation).
for (s in inactivation_systems) stopifnot(kobs(s[[1]], s[[2]], 0) == 0)

# 2. kobs(KI) = kinact/2 exactly (half-maximal at I = KI).
for (s in inactivation_systems) {
  m <- s[[1]]; suf <- s[[2]]
  KI <- ini_val(m, paste0("ki_inact_", suf))
  kinact <- ini_val(m, paste0("kinact_", suf))
  stopifnot(abs(kobs(m, suf, KI) - kinact / 2) < 1e-12)
}

# 3. Plateau: kobs approaches kinact at saturating I.
for (s in inactivation_systems) {
  stopifnot(abs(kobs(s[[1]], s[[2]], 1e7) - ini_val(s[[1]], paste0("kinact_", s[[2]]))) < 1e-5)
}

Known-answer check: the published inhibitory-efficiency ratios

The paper prints the inactivation efficiency kinact / KI for every enzyme system (Discussion), in mL/min/umol. With kinact in /min and KI in uM (= umol/L), kinact / KI is in L/min/umol, i.e. 1000 * kinact / KI is the printed mL/min/umol value. Reproducing all five is a direct check that every kinact and KI pair was transcribed correctly, because an error in either moves the ratio.

efficiency <- function(mod, suffix) {
  1000 * ini_val(mod, paste0("kinact_", suffix)) / ini_val(mod, paste0("ki_inact_", suffix))
}

eff_tbl <- tibble::tibble(
  System = c("SZA pooled CYP3A", "SZA CYP3A4",
             "SZB pooled CYP3A", "SZB CYP3A4", "SZB CYP3A5"),
  `Model kinact/KI (mL/min/umol)` = c(
    efficiency(mod_sza, "3a"), efficiency(mod_sza, "3a4"),
    efficiency(mod_szb, "3a"), efficiency(mod_szb, "3a4"), efficiency(mod_szb, "3a5")
  ),
  `Published (mL/min/umol)` = c(1.856, 1.56, 102.33, 536.2, 18)
)

# The published ratios are rounded; the model recomputes them to full precision.
stopifnot(with(eff_tbl, max(abs(`Model kinact/KI (mL/min/umol)` - `Published (mL/min/umol)`) /
  `Published (mL/min/umol)`) < 0.01))

eff_tbl |>
  dplyr::mutate(`Model kinact/KI (mL/min/umol)` = round(`Model kinact/KI (mL/min/umol)`, 3)) |>
  knitr::kable(caption = "Reproduction of the five published kinact/KI efficiency ratios (He 2022 Discussion).")
Reproduction of the five published kinact/KI efficiency ratios (He 2022 Discussion).
System Model kinact/KI (mL/min/umol) Published (mL/min/umol)
SZA pooled CYP3A 1.856 1.856
SZA CYP3A4 1.560 1.560
SZB pooled CYP3A 102.326 102.330
SZB CYP3A4 536.232 536.200
SZB CYP3A5 18.000 18.000

SZB is one to two orders of magnitude more efficient a CYP3A inactivator than SZA, and SZB is far more efficient on CYP3A4 (536) than on CYP3A5 (18) - the asymmetry the paper attributes to CYP3A5’s higher affinity but lower turnover.

IC50 shift (Results 2.1)

The IC50-shift screen (IC50 without preincubation divided by IC50 after 30-minute NADPH preincubation) is the paper’s first-pass flag for time-dependent inhibition; a value above 1.5 is taken as suggestive. These are reproduced from the printed IC50 values, not carried as model parameters.

ic50_shift <- tibble::tibble(
  Lignan = c("SZA", "SZB"),
  `IC50 no preinc (uM)` = c(63.46, 11.98),
  `IC50 after preinc (uM)` = c(40.45, 0.56),
  `Shift` = c(63.46 / 40.45, 11.98 / 0.56),
  `Published shift` = c(1.57, 21.39)
)
stopifnot(with(ic50_shift, max(abs(Shift - `Published shift`)) < 0.05))
ic50_shift |>
  dplyr::mutate(Shift = round(Shift, 2)) |>
  knitr::kable(caption = "IC50 shift reproduced from the printed IC50 values (He 2022 Results 2.1).")
IC50 shift reproduced from the printed IC50 values (He 2022 Results 2.1).
Lignan IC50 no preinc (uM) IC50 after preinc (uM) Shift Published shift
SZA 63.46 40.45 1.57 1.57
SZB 11.98 0.56 21.39 21.39

Replicating the inactivation time courses (Figures 3A, 4A, 5A, 5C)

The ODE d/dt(E) = -kobs * E with E(0) = 1 has the closed-form solution E(t) = exp(-kobs * t), which is the straight line the authors plot on a log axis. The model is solved over the assay’s preincubation grid at each measured inhibitor concentration.

solve_decay <- function(mod, isoform, conc, covname, tmax = 30) {
  ev <- do.call(rbind, lapply(seq_along(conc), function(i) {
    d <- data.frame(id = i, time = seq(0, tmax, by = 1), evid = 0L, cmt = paste0("enzyme_", isoform))
    d[[covname]] <- conc[i]
    d
  }))
  state <- paste0("enzyme_", isoform)
  out <- rxode2::rxSolve(mod, ev, returnType = "data.frame")
  out$conc <- conc[out$id]
  out$activity <- out[[state]]
  out
}

# SZA on CYP3A4 (Figure 4A): I = 0, 10, 16, 20, 25, 32 uM.
sza_3a4 <- solve_decay(mod_sza, "3a4", c(0, 10, 16, 20, 25, 32), "CP_SZA_UM")
#> Warning: multi-subject simulation without without 'omega'

# SZB on CYP3A4 (Figure 5A): I = 0, 0.5, 1, 2, 4, 8 uM.
szb_3a4 <- solve_decay(mod_szb, "3a4", c(0, 0.5, 1, 2, 4, 8), "CP_SZB_UM")
#> Warning: multi-subject simulation without without 'omega'
# SZB on CYP3A5 (Figure 5C): I = 0, 2, 4, 8, 16 uM.
szb_3a5 <- solve_decay(mod_szb, "3a5", c(0, 2, 4, 8, 16), "CP_SZB_UM")
#> Warning: multi-subject simulation without without 'omega'
# The ODE solution must equal exp(-kobs*t) to solver tolerance.
check_closed_form <- function(df, mod, suffix) {
  df$kobs <- kobs(mod, suffix, df$conc)
  df$analytic <- exp(-df$kobs * df$time)
  max(abs(df$activity - df$analytic))
}
stopifnot(check_closed_form(sza_3a4, mod_sza, "3a4") < 1e-5)
stopifnot(check_closed_form(szb_3a4, mod_szb, "3a4") < 1e-5)
stopifnot(check_closed_form(szb_3a5, mod_szb, "3a5") < 1e-5)
Modelled log-percent CYP activity remaining during preincubation. Replicates the linear-regression lines of He 2022 Figures 4A (SZA on CYP3A4), 5A (SZB on CYP3A4) and 5C (SZB on CYP3A5). SZB's steeper CYP3A4 decay reflects its far larger kinact/KI.

Modelled log-percent CYP activity remaining during preincubation. Replicates the linear-regression lines of He 2022 Figures 4A (SZA on CYP3A4), 5A (SZB on CYP3A4) and 5C (SZB on CYP3A5). SZB’s steeper CYP3A4 decay reflects its far larger kinact/KI.

SZA has no time-dependent inactivation of CYP3A5 (Figure 4C), so enzyme_3a5 in the SZA model holds at its baseline of 1 for all time; that is confirmed directly:

sza_3a5 <- solve_decay(mod_sza, "3a5", c(0, 10, 20, 30, 40, 50), "CP_SZA_UM")
#> Warning: multi-subject simulation without without 'omega'
stopifnot(all(abs(sza_3a5$activity - 1) < 1e-8))

Reversible-inhibition factors

Only SZB is a reversible inhibitor (SZA shows little reversible inhibition, Figure 2A). The Dixon-plot inhibition constants enter a victim-drug clearance as the competitive factor 1/(1 + I/Ki), exposed by the SZB model as riFactor_*. At I = Ki the factor is exactly one half. The two isoform Ki values are close (CYP3A4 2.18 uM, CYP3A5 2.03 uM), with the slightly smaller CYP3A5 Ki making it the marginally stronger reversible target.

ri_factor <- function(I, Ki) 1 / (1 + I / Ki)

# Exact identity at I = Ki for both isoforms and the pooled constant.
for (nm in c("ki_3a4", "ki_3a5", "ki_3a")) {
  ki <- ini_val(mod_szb, nm)
  stopifnot(abs(ri_factor(ki, ki) - 0.5) < 1e-12)
}

# SZB reversibly inhibits CYP3A5 marginally more strongly than CYP3A4: the
# CYP3A5 Ki (2.03 uM) is smaller than the CYP3A4 Ki (2.18 uM), so at a common
# concentration the CYP3A5 activity loss is slightly larger.
I_common <- 2
cyp3a4_loss <- 1 - ri_factor(I_common, ini_val(mod_szb, "ki_3a4"))
cyp3a5_loss <- 1 - ri_factor(I_common, ini_val(mod_szb, "ki_3a5"))
stopifnot(cyp3a5_loss > cyp3a4_loss)
c(SZB_CYP3A4_activity_loss = cyp3a4_loss, SZB_CYP3A5_activity_loss = cyp3a5_loss)
#> SZB_CYP3A4_activity_loss SZB_CYP3A5_activity_loss 
#>                0.4784689                0.4962779

Published results not reproduced here

The Simcyp whole-body half predicted the fold-increase (AUC ratio, AUCR) in tacrolimus exposure when co-dosed with each lignan, split by CYP3A5 genotype and by which inhibition mechanism was switched on (Table 2). In CYP3A5 non-expressers, multiple doses of SZB raised tacrolimus AUC by 57% (AUCR 1.57) under the combined RI+TDI case, while SZA raised it only slightly (AUCR 1.16); the effect was smaller in expressers. These numbers depend on the physicochemical and disposition inputs of Table 3, the tissue-partition method, the organ volumes/flows and enzyme abundances of the Simcyp physiology library, none of which is an nlmixr2 ODE model, so they are recorded here rather than reproduced.

Assumptions and deviations

  • No residual-error model. The source fits the RI and TDI assays by linear and nonlinear regression and reports only the resulting constants; there is no residual-error model and no assay CV. Per the library convention the additive residual SD is fixed at zero, so each model returns the deterministic published curve.
  • No inter-individual variability. The reported +/- terms are standard errors of the regression, not between-donor variability, so they are recorded in population and not encoded as an omega. Pooled and genotyped microsomes are single systems, not populations.
  • Reversible Ki values for SZB. The Discussion swaps the CYP3A4 and CYP3A5 reversible Ki values relative to Results 2.2 and the Figure 2C/2D annotations; the figure/Results order is used here (CYP3A4 Ki = 2.18 uM, CYP3A5 Ki = 2.03 uM).
  • Constants read from figure annotations. The fitted constants are printed in the annotations of Figures 2-5 and repeated in the Results and Discussion text. They are the authors’ own estimates.
  • kinact time unit. kinact is carried in per-minute units, matching the per-minute preincubation-time axis of the assays; the model time unit is minutes.
  • No enzyme turnover. The CYP degradation rate constant kdeg, which governs the in-vivo magnitude of TDI, is a Simcyp default and is not printed. No turnover term is carried, so the models describe the inactivation phase of the in-vitro assay only; resynthesis is negligible over the 30-minute incubation.
  • Mechanisms carried per isoform as fitted. SZA has no reversible inhibition (Figure 2A) and no CYP3A5 inactivation (Figure 4C), so that model carries only the CYP3A4 (and pooled) inactivation terms and enzyme_3a5 holds at baseline. SZB inhibits both isoforms by both mechanisms. Each file also carries the pooled-HLM CYP3A screen as ini() constants exposed through derived outputs.
  • In-vivo use requires an unbound driving concentration. Applying either model to a plasma profile requires supplying the unbound inhibitor concentration at the site of enzyme interaction as CP_SZA_UM / CP_SZB_UM; the in-vitro incubation concentration is not a plasma concentration. ```