The plant and the material

Dendrobium officinale Kimura et Migo is an orchid with fleshy, jointed stems. The dried stem, often twisted into spirals and sold as fengdou (枫斗), is the material known in Chinese as tiepi shihu (铁皮石斛), “iron-skin Dendrobium”. The Chinese Pharmacopoeia lists it as a separate monograph from other Dendrobium species, with its own identity tests and assays (国家药典委员会 2020).

A large share of the stem’s dry weight, a quarter or more in material that meets the pharmacopoeia’s minimum, is a water-soluble polysaccharide. This is the molecule that the pharmacopoeia measures and that the market talks about, so it is worth knowing exactly what it is.

A glucomannan with acetyl groups

A polysaccharide is a long chain of sugar units. In Dendrobium officinale the main chain is made of mannose and glucose, both six-carbon sugars, joined by β-(1→4) links, the same kind of link that holds the glucose units of cellulose together. A chain of mannose and glucose like this is called a glucomannan. Konjac glucomannan, from the corm of Amorphophallus konjac, is a familiar relative.

Three features matter for quality. The first is the mannose-to-glucose ratio, which identifies the material and distinguishes it from other glucomannans and from starch, which is all glucose. The second is the acetyl groups: small CH$_3$CO– groups attached to the sugar units, which stop the chains packing tightly together and so help keep the polysaccharide soluble in water. The third is molecular weight, which affects viscosity and, according to many laboratory studies, biological activity. All three can change with the plant’s age and growing conditions, and with how the stems are dried and how the polysaccharide is extracted.

# 1. Mannose and glucose after acid hydrolysis and PMP labelling (HPLC).
#    Peak areas and response factors are invented for the example.
area = {"mannose": 41.6, "glucose": 9.1}            # sample peak areas
resp = {"mannose": 1.00, "glucose": 1.07}           # area per µg, from standards
ug = {s: area[s] / resp[s] for s in area}
print(f"peak-area ratio Man/Glc {area['mannose'] / area['glucose']:.2f} "
      f"(window 2.4–8.0); mass ratio {ug['mannose'] / ug['glucose']:.2f}")

# 2. Published molar ratios differ widely between samples and methods:
for source, man, glc in (("Hua et al. 2004", 40.2, 8.4), ("Guo et al. 2022", 10.6, 1.0)):
    print(f"  {source}: Man/Glc = {man / glc:.1f}")

# 3. Degree of acetylation from the acetyl content by weight.
#    Each hexose residue weighs 162 g/mol; an acetyl group adds 42 g/mol
#    (CH3CO, 43, replacing one H, 1). w = mass fraction of acetyl (CH3CO).
def degree_of_substitution(w):
    return 162 * w / (43 - 42 * w)

for w in (0.0684, 0.10):
    print(f"acetyl {100 * w:.2f} % by weight → "
          f"{degree_of_substitution(w):.2f} acetyl groups per sugar residue")
# peak-area ratio Man/Glc 4.57 (window 2.4–8.0); mass ratio 4.89
#   Hua et al. 2004: Man/Glc = 4.8
#   Guo et al. 2022: Man/Glc = 10.6
# acetyl 6.84 % by weight → 0.28 acetyl groups per sugar residue
# acetyl 10.00 % by weight → 0.42 acetyl groups per sugar residue

The two published ratios differ by more than a factor of two. That is typical: different samples, different extraction and purification steps, and different analytical methods give different numbers. A purified fraction studied in a laboratory is not the same thing as the crude polysaccharide in a whole stem. It is one reason the pharmacopoeia specifies its own methods and a wide acceptance window rather than a single target.

The pharmacopoeia assays

The Chinese Pharmacopoeia monograph for Dendrobium officinale sets, among other tests, the following limits on the dried material (国家药典委员会 2020, 2021):

The polysaccharide assay uses a reaction described by Dubois and colleagues in 1956 (Dubois et al. 1956). Concentrated sulfuric acid dehydrates sugar units to furan derivatives, which react with phenol to give an orange-yellow colour. The intensity of the colour is measured with a spectrophotometer and converted to an amount of sugar with a standard curve.

The Beer–Lambert law and the standard curve

The conversion rests on the Beer–Lambert law. For a coloured solution, the absorbance $A$ (the logarithm of the ratio of incoming to transmitted light) is proportional to the concentration $c$ of the coloured substance, its molar absorptivity $\varepsilon$ and the path length $l$ of the cell:

$$A = \varepsilon\, l\, c$$

In practice $\varepsilon$ is not calculated; it is measured, by running known amounts of a reference substance through the same reaction and fitting a straight line. The pharmacopoeia method uses anhydrous glucose at 90 µg per mL, in volumes from 0.2 to 1.0 mL, so the standard tubes contain 18 to 90 µg of glucose.

from statistics import mean

# Standard curve: anhydrous glucose, 90 µg/mL, 0.2 to 1.0 mL per tube,
# phenol–sulfuric acid colour, absorbance at 488 nm.
# The absorbance readings are invented for the example.
glucose_ug = [90 * v for v in (0.2, 0.4, 0.6, 0.8, 1.0)]
absorb = [0.142, 0.281, 0.425, 0.560, 0.706]

mx, my = mean(glucose_ug), mean(absorb)
sxy = sum((x - mx) * (y - my) for x, y in zip(glucose_ug, absorb))
sxx = sum((x - mx) ** 2 for x in glucose_ug)
syy = sum((y - my) ** 2 for y in absorb)
slope = sxy / sxx
intercept = my - slope * mx
r = sxy / (sxx * syy) ** 0.5
print(f"A = {intercept:.4f} + {slope:.5f} × (µg glucose),  r = {r:.4f}")

# One sample (invented): 0.3000 g powder, extract made up to 100 mL,
# 5 mL of extract diluted to 100 mL, 1 mL of that taken for colour.
a_sample = 0.355
ug_in_tube = (a_sample - intercept) / slope
dilution = (100 / 1) * (100 / 5)        # µg in tube → µg in whole extract
mg_total = ug_in_tube * dilution / 1000
moisture = 0.098                          # 9.8 % water, from a separate test
pct_as_is = 100 * mg_total / 300.0
pct_dry = pct_as_is / (1 - moisture)
print(f"µg in tube {ug_in_tube:.1f}; polysaccharide as glucose: "
      f"{pct_as_is:.1f} % as is, {pct_dry:.1f} % dry basis")
print("meets the 25.0 % minimum (dry basis):", pct_dry >= 25.0)
# A = 0.0007 + 0.00782 × (µg glucose),  r = 0.9999
# µg in tube 45.3; polysaccharide as glucose: 30.2 % as is, 33.5 % dry basis
# meets the 25.0 % minimum (dry basis): True

What the numbers do not tell you

A stem that passes every pharmacopoeia test is genuine Dendrobium officinale of acceptable quality by that standard. It is not thereby shown to have any health effect in people. The many laboratory and animal studies on Dendrobium officinale polysaccharides, on immune cells, gut bacteria, blood sugar and more, are interesting, but they are not evidence that eating the stems or drinking an extract prevents or treats any human disease, and this book makes no such claim.